Embed Size (px)
Citation preview
PERSONAL INFORMATION MANAGEMENT
AMONG OFFICE SUPPORT STAFF IN A UNIVERSITY ENVIRONMENT:
AN EXPLORATORY STUDY
François-Xavier Paré
School of Information Studies
McGill University, Montreal
June 2011
A thesis submitted in partial fulfillment of the requirements of the degree of
Doctor of Philosophy (Information Studies)
© François-Xavier Paré, 2011
Table of contents
Abstract ........................................................................................................................... v
Résumé .......................................................................................................................... vii
Acknowledgements ........................................................................................................ ix
Chapter 1: Introduction ................................................................................................... 1
1.1 Purpose of this study ............................................................................................. 2
Chapter 2: Literature Review and Conceptual Framework ............................................ 6
2.1 Defining Personal Information Management ....................................................... 7
2.2 Related disciplines ............................................................................................... 8
2.2.1 Information behaviour .................................................................................. 9
2.2.2 Psychology and cognitive science .............................................................. 10
2.2.3 Knowledge organization/Classification ...................................................... 11
2.2.4 Human-computer interaction ...................................................................... 12
2.2.5 Ethnography ................................................................................................ 14
2.2.6 Knowledge Management ............................................................................ 15
2.3 PIM Studies ........................................................................................................ 16
2.3.1 Early studies ................................................................................................ 16
2.3.2 The 1980s onward: the natural history of offices and the importance of
context ................................................................................................................... 17
2.3.3 PIM in the digital environment ................................................................... 20
2.4 Context of the present study .............................................................................. 25
2.5 Information, documents and files: terminological considerations ..................... 27
2.5.1 Information ................................................................................................. 28
2.5.2 Documents .................................................................................................. 28
2.5.3 Files ............................................................................................................. 29
2.6 Conceptual framework ....................................................................................... 30
2.6.1 Context of document organization .............................................................. 31
2.6.2 Job content .................................................................................................. 32
i
2.6.3 Document attributes .................................................................................... 32
2.6.4 Individual preferences and abilities ............................................................ 33
2.6.5 Towards a conceptual model ...................................................................... 33
Chapter 3: Methodology ............................................................................................... 35
3.1 Grounded Theory ............................................................................................... 35
3.2 Data Collection and analysis .............................................................................. 36
3.2.1 Pilot study ................................................................................................... 40
3.2.2 Ethics considerations .................................................................................. 41
3.3 Setting and sample ............................................................................................. 42
3.4 Validity and trustworthiness in grounded theory ............................................... 56
Chapter 4: Survey of document spaces ......................................................................... 61
4.1 Defining document spaces ................................................................................. 61
4.2 Windows file structure ....................................................................................... 62
4.2.1 Main folder.................................................................................................. 63
4.2.1.1 Characteristics of the file folder structure ............................................ 64
4.2.1.2 Main organizational criteria ................................................................. 67
4.2.1.3 Orphan Files ......................................................................................... 68
4.2.1.4 Dealing with multiple versions ............................................................ 71
4.2.1.5 Naming files and folders ...................................................................... 72
4.2.1.6 Searching and browsing ....................................................................... 74
4.2.2 Secondary folders........................................................................................ 76
4.2.3 Windows Desktop ....................................................................................... 81
4.2.3.1 Templates and reference documents .................................................... 81
4.2.3.2 Documents currently being worked on ................................................ 82
4.2.3.3 Documents in transit ............................................................................ 84
4.2.3.4 Reminding ............................................................................................ 85
4.2.3.5 Messiness ............................................................................................. 86
4.3 E-mail ................................................................................................................. 86
4.3.1 Weeding ...................................................................................................... 87
ii
4.3.2 Attachments ................................................................................................ 88
4.3.3 Correspondence between Windows folders and e-mail .............................. 90
4.4 Paper documents ................................................................................................ 91
4.5 Shared environments .......................................................................................... 95
4.5.1 Coordination mechanisms ........................................................................... 99
Chapter 5: Factors and causes of variations ................................................................ 101
5.1 Job Content ...................................................................................................... 102
5.1.1 Breadth and complexity ............................................................................ 102
5.1.2 Work Pace ................................................................................................. 104
5.1.3 Cyclical nature of the work ....................................................................... 106
5.1.4 Domain and context .................................................................................. 107
5.2 Job status .......................................................................................................... 109
5.3 Existing documents .......................................................................................... 112
5.3.1 Help from the predecessor ........................................................................ 114
5.3.2 Appropriation strategy .............................................................................. 117
5.4 Relationship with the superior ......................................................................... 120
5.4.1 Nature of the working relationship ........................................................... 121
5.4.2 Time working together .............................................................................. 122
5.4.3 Personality of the superior ........................................................................ 123
5.4.4 Change of superior .................................................................................... 126
5.5 Worker characteristics ..................................................................................... 126
5.5.1 Personality traits ........................................................................................ 127
5.5.2 Preferences and dislikes ............................................................................ 129
5.5.3 Training and education ............................................................................. 130
5.5.4 Experience................................................................................................. 130
5.5.5 Computer literacy...................................................................................... 133
5.5.6 Cognitive style and abilities ...................................................................... 135
5.6 Organizational context ..................................................................................... 135
5.7 Document attributes ......................................................................................... 138
5.8 Development of a conceptual model ................................................................ 142
iii
5.9 Limitations of the study ................................................................................... 144
Chapter 6: Conclusion ................................................................................................ 146
6.1 Summary of the findings .................................................................................. 146
6.2 Significance of this study ................................................................................. 149
6.3 Future research and recommendations for researchers .................................... 151
6.4 Recommendations for practitioners ................................................................. 152
Bibliography ............................................................................................................... 155
Appendix 1: Example of an e-mail invitation to participate in the study ................... 172
Appendix 2: Interview Guide ..................................................................................... 173
Appendix 3: Pre-interview survey .............................................................................. 175
Appendix 4: Participant profiles ................................................................................. 188
Participant #1: Anne ............................................................................................... 189
Participant #2: Béatrice ........................................................................................... 190
Participant #3: Carole ............................................................................................. 191
Participant #4: Diane ............................................................................................... 192
Participant #5: Édith ............................................................................................... 193
Participant #6: Francine .......................................................................................... 194
Participant #7: Ginette ............................................................................................ 195
Participant #8: Hortense .......................................................................................... 196
Participant #9: Ingrid .............................................................................................. 197
Participant #10: Jeanne ........................................................................................... 198
Participant #11: Kim ............................................................................................... 199
Participant #12: Lise ............................................................................................... 200
Participant #13: Manon ........................................................................................... 201
Participant #14: Nicole ........................................................................................... 202
Participant #15: Olivia ............................................................................................ 203
iv
Abstract
Since the late 1960s, several studies have investigated personal information management
(PIM) in the workplace. However, very few studies have focused on the behaviour of
office support staff in a work environment. The purpose of this exploratory study was to
examine the document management behaviour of office support staff in a large Canadian
university.
The methodological approach used for this study was grounded theory. Fifteen in-depth
interviews were conducted in participants' offices, and visual observations of their
document structures were made. A pre-interview survey was also administered in order to
gather additional information. Participants were chosen according to the principles of
theoretical sampling, and simultaneous data collection and analysis continued until
theoretical saturation was reached. Transcribed interviews were coded, after which
abstract concepts were derived and grouped into categories, using the constant
comparison method. A substantive theory was then developed.
The findings suggest the existence of several distinct document spaces within workers'
document landscape: a main folder, secondary folders, the operating system desktop, e-
mail, paper documents and shared environments. Behaviour pertaining to the handling of
orphan files and multiple versions, the naming of files and folders as well as searching
and browsing were described. Overall, despite several elements in common, significant
variation was noted among participants.
In order to explain the variation observed, a model of the factors that are likely to
influence PIM behaviour was developed. It comprises seven main categories of factors:
job content, job status, existing documents, relationship with the superior, worker
characteristics, organizational context and document attributes. Several of the factors
identified had never been mentioned in the PIM literature, while in other cases, the
evidence presented helped confirm previous findings. The proposed model also
v
highlights the inherent complexity of PIM, and the importance of adopting an all-
encompassing view when analyzing PIM behaviour.
vi
Résumé
Depuis la fin des années 1960, plusieurs études ont porté sur la gestion personnelle de
l'information (GPI) en milieu de travail. Néanmoins, très peu d’études se sont penchées
sur les comportements des employés de soutien tels que les commis de bureau ou les
assistantes administratives. L'objectif de cette étude exploratoire était d'examiner les
comportements de gestion documentaire d'employés de soutien dans une grande
université canadienne.
L'approche méthodologique utilisée est celle de la théorisation ancrée. Quinze entrevues
en profondeur ont été menées dans le bureau des participantes, et des observations
visuelles de leurs structures de documents ont été effectuées. Un questionnaire pré-
entrevue a également permis de recueillir des informations additionnelles. Les
participants ont été sélectionnés selon un principe d'échantillonnage théorique, et la
collecte et l’analyse des données, menées en parallèle, se sont poursuivies jusqu'au point
de saturation théorique. Les entrevues transcrites ont été codées, à la suite de quoi des
concepts ont été dérivés et groupés en catégories, selon la méthode de la comparaison
constante. Une théorie substantive a ensuite été développée.
Les résultats suggèrent que le paysage documentaire de ces employés est composé de
plusieurs espaces documentaires distincts: un répertoire principal, des répertoires
secondaires, le bureau du système d'exploitation, le courriel, les documents papiers ainsi
que les environnements partagés. Les habitudes des participants face à la gestion des
fichiers orphelins, des versions multiples, au nommage des fichiers et répertoires ainsi
qu'à la recherche et la navigation ont été décrites. En somme, malgré certains éléments en
commun, des variations significatives ont pu être observées parmi les participants.
Afin d'expliquer la variation observée, un modèle des facteurs pouvant influencer la GPI
a été développé. Il comprend sept catégories de facteurs: nature de l'emploi, statut de
l'emploi, documents existants, relation avec le ou la supérieur(e), caractéristiques de
l'employé, contexte organisationnel et caractéristiques des documents. Plusieurs des
vii
facteurs identifiés dans cette étude n'avaient jamais été mentionnés dans la littérature sur
la GPI, alors que dans d'autres cas les éléments présentés ici viennent confirmer les
résultats d'études précédentes. Le modèle proposé met également en lumière la
complexité inhérente à la GPI, et l'importance d'adopter une approche holistique dans
l'analyse des comportements de GPI.
viii
Acknowledgements
Many people have played a role in one way or another in the completion of this
dissertation - so many in fact that I inevitably run the risk of forgetting someone. If that is
the case, please accept my preemptive apologies and sincere gratitude.
I would first and foremost like to thank my advisor, Professor Jamshid Beheshti, for his
availability, patience and understanding throughout this long journey. His wisdom and
presence of mind have no doubt helped me avoid many a pitfall.
Sincere thanks also go out to:
The members of my doctoral committee, Professors Andrew Large, John Leide
and Doreen Starke-Meyerring, for their insightful comments and suggestions;
Professor Éric Leroux, for his continuing friendship and guidance;
Professor Sabine Mas, for enthusiastically allowing me to use the survey she had
developed for her own dissertation;
Professors Kimiz Dalkir, Judith Lavoie, Christine Dufour, Leanne Bowler,
Valerie Nesset, Audrey Laplante, Clément Arsenault, James Turner and Carol
Couture, who have helped me in various ways, whether they realized it or not;
My friends, Professors Lei Zu (祖垒) and Jin Zhang (张锦), without whom I
probably wouldn't have undertaken doctoral studies; and Mr. Lionel Villalonga
for his precious help;
The staff at McGill's School of Information Studies for their warm welcome on
my all-too-rare visits;
ix
x
My coworkers at Université de Montréal, especially Paul-Émil Provost for his
interest in my research and Carole Paradis for dealing with my somewhat
unconventional vacation schedule.
Although I am not at liberty to reveal their names here, I would like to convey my
deepest gratitude to all the volunteers who participated in this study for their precious
time and efforts, and to their supervisors for allowing me to conduct these interviews.
I would also like to take this opportunity to thank my parents for everything they've ever
done for me. They, more than anybody else, have made me what I am today, and for that
I am grateful.
And last but not least, sincere thanks to Xiaoyan Xie (谢晓燕) for her love and tireless
encouragement.
Chapter 1: Introduction
Perhaps now more than ever, efficient document management is crucial to any
organization’s success (Eppler & Mengis, 2003). In today’s organizations, employees
often have to deal with significant amounts of information, and managing that
information is now an unavoidable component of many office workers’ daily tasks.
Through several studies, it has been estimated that office workers can spend 25 to 60
percent of their time working with documents (Gordon, 1997; Meier & Sprague, 1996).
The efficient management of information is therefore a major factor in the overall
productivity of office workers.
Despite the implementation of complex, large-scale knowledge and document
management systems, a significant part of an organization’s documents still resides on its
employees’ computers or in other environments over which they have full control, such
as shared folders. These documents form a kind of personal information collection, which
can be seen as a personalized subset of all information available (Bruce, 2005).
As a result, office workers often enjoy a significant degree of freedom when it comes to
managing their documents. Consequently, every worker tends to develop his/her own
particular methods and habits for handling the electronic documents under their control.
The study of Personal Information Management (PIM) aims to investigate those
behaviours, and eventually develop tools to enhance the management of information
(Barreau & Nardi, 1995; Landsdale, 1988; Neumann, 1989). It has been defined as “the
practice of managing the information that helps us in our daily lives”, which of course
encompasses work situations (Bellotti & Smith, 2000, p. 227). The word personal in PIM
alludes to the fact that the information is used and under direct control of an individual,
although the information is not necessarily private, confidential, or even about the
individual (Henderson, 2005, Landsdale, 1988).
In the last forty years or so, PIM researchers have tried to better understand individual
office workers’ information management behaviours. As a result, there is now a
1
significant and growing body of knowledge exploring the way these workers handle
information in the workplace.
1.1 Purpose of this study
The present study seeks to add to this body of knowledge by looking more specifically at
the document management behaviours of a specific type of worker: office support staff.
We use the term office support staff to refer to employees whose work involves
supporting the daily activities of an organization. Typical positions include office clerks,
office technicians and administrative assistants, traditionally called secretaries1.
This type of worker is extremely widespread in organizations around the world.
According to the U. S. Bureau of Labor Statistics 2 , there were about 4.3 million
secretaries and administrative assistants in the United States in 2008. The IAAP estimates
the number at over 500 000 for Canada in 20113.
The exact tasks performed by office support staff tend to vary considerably according to
the needs of each organization. Nevertheless, according to the U.S. Bureau of Labor
Statistics,
"[s]ecretaries and administrative assistants perform a variety of administrative
and clerical duties necessary to run an organization efficiently. They serve as
1 A brief note on terminology. The term secretary is now somewhat frowned upon (Durst, 1994), and has
been progressively abandoned in recent years, notably by the National Secretaries Association. This
association, which was founded in 1942, changed its name to the International Association of
Administrative Professionals (IAAP) in 1998. That being said, secretary still appears to be widely used in
the workplace and in common usage. In fact, some of the participants in this study referred to themselves
secretaries, even if that was not their official job title. 2 http://data.bls.gov/cgi-bin/print.pl/oco/ocos151.htm 3 http://www.iaap-hq.org/events/apw/fastfacts
2
information and communication managers for an office; plan and schedule
meetings and appointments; organize and maintain paper and electronic files;
manage projects; conduct research; and disseminate information by using the
telephone, mail services, Web sites, and e-mail. They may also handle travel
and guest arrangements." (U.S. Bureau of Labor Statistics, 2009)
There are several reasons why this type of worker is especially interesting in the context
of PIM. First, support staff forms an essential part of any organization, yet there has been
very little attention given to the information behaviour of support staff in organizations,
at least from a PIM perspective4. In fact, the bulk of the PIM literature has repeatedly
focused on a select few types of individuals: professors/researchers (Case, 1986; Kwasnik,
1991), professionals (Jones et al., 2002) and managers (Barreau, 1995; Mackenzie, 2000),
who could all be considered so-called knowledge workers (Drucker, 1973; Kidd, 1994).
In fact, some have argued that knowledge workers make more interesting participants in
PIM studies, because of the supposedly non-linear nature of their work (Kidd, 1994;
Malone, 1983). Knowledge workers are traditionally perceived as having more complex
and fractured tasks, which may seem more interesting to analyze. But while it may be
true that the nature of office support staff’s work is sometimes more clerical, it is not
necessarily devoid of complexity. In fact, an administrative assistant’s information space
may well be as complex as his/her superior’s, since they are likely to work on the very
same projects, albeit of course at different levels.
Another reason why office support staff may be good candidates for PIM studies is that a
significant part of their tasks precisely involves working with documents. Indeed, the U.S.
Bureau of Labor Statistics (2009) summarizes administrative assistants’ responsibilities
4 Sabine Mas’ (2007) dissertation is a notable exception. The participants for the first half of her study were
office support staff workers in an academic environment. However, her study focused on the quantitative
aspects of workers’ file structures and their impact on information retrieval. Relatively little was said about
this type of workers’ specific PIM-related habits and preferences.
3
as “performing and coordinating an office's administrative activities and storing,
retrieving, and integrating information for dissemination to staff and clients.” As can be
seen from this description, information is a central element to their work5.
One more interesting aspect of assistants’ work, for instance, is that on the one hand they
may work in a relatively structured frame, with rules set either by their superior or by
their department or unit, but on the other hand they may also enjoy a certain degree of
freedom with regard to how they organize their documents. This apparent contradiction
may not be as common for other types of workers, which makes support staff especially
interesting to study.
The aim of this study is thus to learn more about the specificities of document
management as done by office support staff. To that end, we shall focus on the following
research questions:
How do office support staff organize their documents?
Why do office support staff organize their documents the way they do?
These questions are relatively broad, but they are intentionally so. Given that very few
studies have so far been devoted to this type of worker, we are treading relatively
uncharted territory. As such, a broader view of the problem was deemed more
appropriate. The present study is thus rather exploratory in nature; subsequent studies
may eventually focus on more specific aspects.
Therefore, our intent is mainly to observe how workers handle and organize documents in
the context of their work, and examine the behaviours they have adopted, the specific
habits they have developed, and the document structures they have created. This applies
to documents in both digital and paper form. We then intend to find out why PIM tasks
5 The distinction between information, documents and files in the context of this study will be discussed in
Chapter 2.
4
are done a certain way, and try to understand the motivations or structural conditions that
may influence workers’ behaviours.
The following chapter presents a literature review on Personal Information Management
and pertinent topics as related to the two research questions, and explains the conceptual
framework of the research. Its purpose is to introduce some of the major studies in the
field, and thus serve as a background to our own study.
Chapter 3 contains a detailed accounts of the methods used for data collection and
analysis. Chapters 4 and 5 will focus on a presentation and an analysis of the data
collected. We will then conclude in Chapter 6.
5
Chapter 2: Literature Review and Conceptual Framework
In his famous 1945 article entitled “As we may think”, Vannevar Bush envisioned an
imaginary apparatus called the memex, which he described as:
“(…) a device in which an individual stores all his books, records, and
communications, and which is mechanized so that it may be consulted
with exceeding speed and flexibility. It is an enlarged intimate supplement
to his memory.” (Bush, 1945).
While it is true that today’s technology makes Bush’s vision closer to reality, there is still
a long way to go. Despite having ever more powerful tools at their fingertips, people still
routinely experience problems in managing information, whether in their work or in their
daily life (Bellotti & Smith, 2000; Burton, 1981; Landsdale, 1988; Nardi & Barreau,
1997; Neumann, 1989). The problem is somewhat exacerbated by the sheer amount
information that is available to us nowadays, and that we must process in order to
accomplish our work, or simply function in society. There is increasing recognition that
information overload is a genuine problem, and that its consequences are of prime
concern for organizations (Edmunds & Morris, 2000).
The growing importance of information, and the need to manage it adequately, have led
to the birth of an area of study called Personal Information Management (PIM), which
seeks to examine how individuals manage information in their daily lives, notably in the
work environment. PIM can be considered a branch of information studies, although it
tends to borrow extensively from other disciplines as well, both for its theoretical
foundations and its methodologies.
It could be argued that by describing the memex more than 65 years ago, Vannevar Bush
was setting the future research agenda for PIM. And in many ways, the memex still stands
as the Holy Grail for a number of PIM researchers and developers, who try to develop a
ubiquitous device that would make document management easy for everyone. But PIM’s
6
objectives go deeper than simply achieving the memex; it strives to understand how
people make sense of, and manage their documents, something that would probably still
be necessary in a memex-world.
In the last 40 years or so, much has been discovered about how people manage, classify,
file and retrieve their documents. This chapter presents an overview of these discoveries,
via a review of the literature in PIM. We first define PIM and examine how its
development and theoretical foundations were influenced by other disciplines such as
information behaviour, psychology, knowledge organization, human computer interaction
and ethnography. We then look at some of the most important research done in the area
of PIM over the years: the early information use studies made in the 1960s and 1970s, the
workplace-influenced studies of the early 1980s, and finally, PIM in the digital
environment. A section on the distinction between information, documents and files
follows, after which we present a basic conceptual framework for our study, based on the
literature reviewed.
2.1 Defining Personal Information Management
As early as 1960, Engelbart dubbed the management of personal documentation
microdocumentation, as opposed to the macrodocumentation, which consisted of libraries
and related systems. Twenty years later, Hilary Burton defined personal information
systems as involving “the acquisition, storage, classification and/or retrieval of
information by an individual for himself” (Burton, 1980, 441). In many respects, this
definition is still valid today; although it is fairly broad and quite basic, it adequately
summarizes the concerns of researchers in the field to this day. The expression personal
information management hadn’t yet been coined at the time this definition was written,
but all the important conceptual elements were already present.
And indeed not much was added to that definition in the following years, although there
was a definite shift in emphasis. Fifteen years after Burton, Barreau (1995) talked about a
personal information management system as “an information system developed by or
7
created for an individual for personal use in a work environment”. Such a system would
include:
“a person’s methods and rules for acquiring the information which
becomes part of the system, the mechanisms for organizing and storing
the information, the rules and procedures for maintaining the system, the
mechanisms for retrieval, and the procedures for producing the various
outputs required” (Barreau, 1995, 327).
At its core, this view is quite similar to Burton’s. The main difference is the emphasis on
the workplace as the environment for PIM. As we shall see below, this reflects the fact
that most of the PIM studies conducted in the 1980s, 1990s and 2000s were done in a
work context, most often studying employees and managers with different levels of
responsibility. The idea that systems could be developed for an individual is also
indicative of the numerous systems developed in that period to help support work-related
PIM.
So what exactly constitutes the personal information we need to manage? Bellotti and
Smith mention “addresses, phone numbers, to-dos, appointments, notes, documents,
folders […] urls [sic] and email addresses” (Bellotti & Smith, 2000, 227). In other words,
we may consider that just about any piece of information which passes through a person’s
hands becomes personal information, and thus has to be handled - managed - by that
person. Although some studies have focused on a single type of document, the scope of
PIM covers all of the information a person might possess, whatever the format.
2.2 Related disciplines
Though grounded in information studies6, personal information management stands at the
crossroads of several disciplines. The first PIM studies were of a very exploratory nature,
6 In this study, the terms information science and information studies are used interchangeably.
8
simply trying to observe how people organized their offices. Researchers then started
looking elsewhere to make sense of the data they had obtained. In this section we shall
briefly examine some of the disciplines PIM researchers have tapped into over the years.
2.2.1 Information behaviour
PIM has strong connections with, and in many ways originates from, the area of
information behaviour. The first researchers that touched upon PIM were in fact studying
the information behaviour of scientists, more specifically their personal indexes and
collections of documents (Jahoda et al., 1966 ; Soper, 1976). Those early studies were
often more concerned with the amount of documentation present in their subject’s offices
– number of books, articles, etc. – than its organization, although exceptions such as
Engelbart (1960) or Murphy (1980) did focus on the organizational methods of their
subjects rather than the amount of materials they owned. Nevertheless, the interest in the
use and organization of the information was always present in these studies, and this
would eventually become the main concern of PIM.
It could be argued that PIM is actually a part of information behaviour (IB), since IB
involves “the study of how people need, seek, give, and use information in different
contexts, including the workplace and everyday living” (Pettigrew et al., 1999). Yet in
reality there doesn’t seem to be a big intersection between the two areas, and one would
be hard-pressed to find several researchers who work in both domains. It seems that the
focus in IB is more on the user’s information needs, and the ways to satisfy them (Wilson,
1999), whereas PIM is more concerned with people keeping, classifying and organizing
their own documents. Whether they are separated or not, it seems safe to assert that these
two areas of interest have a lot to learn from each other, and that they could eventually be
much closer than they are at this point.
9
2.2.2 Psychology and cognitive science
Some researchers tend to emphasize the psychological aspect of PIM, and the theory in
psychology and cognitive science has often informed PIM experiments. Lansdale (1988)
identified the three main processes involved in information management as categorization,
recall and recognition.
Categorization by human subjects has long been the object of research in psychology, and
concepts such as cognitive economy, perceived world structure and the appropriateness
of prototypes (Rosch, 1978) have been helpful in understanding how people classify
objects and ideas. Recall and recognition also play a crucial role in managing information,
since, as Landsdale (1988, 55) reminds us: “The primary reason […] for keeping this
information is to be able to retrieve and use it in the future.” In other words, after
categorizing an item, one has to remember what label was assigned to it. Thus, the link
between organization of information and memory is of primary importance (Najarian,
1978), and over the years, PIM researchers have tried to identify how documents could
best be remembered. Factors such as general appearance, color, format, location of the
document (Landsdale, 1991), context (Kwasnik, 1991) or chronological events (Ringel et
al., 2003) have all been studied in the hopes of identifying the best possible reminders for
documents. According to Parker and Paisley, the ideal PIM system would simply be an
extension of the user’s own memory, and could be queried in a similar manner, echoing
Vannevar Bush’s vision (Parker & Paisley, 1966).
It has been shown that interacting with an interface, as is the case in all computer-based
PIM, involves cognitive abilities such as perception, recognition and attention (Alkhalifa,
2006). Cognitive load is considered an important factor in navigating a hierarchical file
structure (Boechler, 2006; Jones et al., 2005). PIM tasks can also require problem-solving
and decision-making skills (Jones, 2008). In a recent survey of the literature, Gwizdka &
Chignell (2007) identified several cognitive and affective factors which are likely to play
a role in PIM, among which figure flexibility of closure, working memory, spatial ability,
spatial scanning and cognitive style.
10
Parker and Paisley (1966) once warned that systems that were being built for users would
inevitably have to deal with a certain level of apparent irrationality and inefficiency on
the part of users. Just because a system is conceived in a supposedly rational way, it
doesn’t mean it will be consistent with the way users think. In light of this and all of the
above, it is safe to say that much more research on the psychological and cognitive
aspects of PIM is needed.
2.2.3 Knowledge organization/Classification
Knowledge organization (KO) and classification theory are another source of theory for
PIM, because they provide principles to help analyze the structure and effectiveness of
human-produced classifications and categorizations. User-developed PIM systems share
several characteristics with traditional bibliographic systems, notably in their objectives:
a PIM system should also, to some extent, help in finding, collocating, choice, acquisition
and navigation (Svenonius, 2000). Therefore, criteria developed to evaluate bibliographic
systems could potentially be used to evaluate PIM systems. Related domains such as
ontologies and taxonomies are also of some importance in PIM (McIllwaine, 2003).
The developments in automatic classification could also play a major role in PIM,
although this is the object of debate within the field (Mackenzie, 2002; Whittaker &
Sidner, 1996). But even if computers might not be able to replace humans in all
classification tasks, many agree that automatic classification could at least serve to
facilitate certain tasks for the users (Boardman, 2001b; Figuerola et al., 2001). And if KO
can be useful in the realm of PIM, the opposite might also be true, since, according to
Hjorland: “In general, our knowledge of how humans classify is limited” (Hjorland, 2003,
104). Although we now have a significant body of work related to people’s
organizational behaviours and the problems they face in such tasks, there is still a lot of
uncharted territory. Additional PIM studies could help address this.
11
2.2.4 Human-computer interaction
The arrival of personal computers in the workplace signaled a new era for PIM research.
Given the significant increase in the proportion of documents in digital form, it was
necessary to verify if what had been found in a paper-based environment remained true in
the digital realm (Barreau, 1995). Furthermore, in their early years, computers had mainly
been used by programmers and other expert users. But with the democratization of these
tools, it was also necessary to ensure that regular people could use them with relative
ease, and in a productive manner (Borgman, 1984). In that respect, PIM and computer
science, more specifically human-computer interaction (HCI) share a common interest.
HCI concerns itself with “the design, evaluation and implementation of interactive
computing systems for human use in a social context, and with the study of major
phenomena surrounding them” (Ghaoui, 2006, p. xiv). Since a large part of PIM tasks are
done using a computer, any insights or innovations in HCI can potentially be beneficial
for PIM as well. A good example is User-Centered Design, a design methodology in
which the end-user is placed at the center of the process (Nelson, 2002). The User-
Subjective Approach is an example of an HCI-inspired approach which specifically
addresses PIM concerns (Bergman, Beyth-Marom & Nachmias, 2008).
The most concrete examples of the convergence of HCI and PIM are the multitude of
PIM prototypes that have been developed through the years. In fact, as early as the 1960s,
tools were being developed specifically to help people manage their information, mainly
bibliographic information. Examples include SURF, FAMULUS and GIPSY (Burton,
1981). With the growing importance of personal computers in the 1980s and 1990s, the
HCI community became more prominently involved in PIM research, which yielded
several interesting prototypes. Some of these were enhancements to current Windows-
type operating systems (Gwizdka, 2002; Yiu, 1997), while others were entirely new
systems or environments. A thorough overview of past and current PIM prototypes is
beyond the scope of this paper. Nevertheless, we shall briefly discuss some of the most
important PIM prototypes.
12
First, several researchers have tried to revise the traditional setup of the folder hierarchy.
Projects like Lifestreams (Fertig et al., 1996b), Presto (Dourish et al., 1999) and Haystack
(Karger et al., 2003) moved away from having documents in a fixed location in the file
system. For instance, Presto uses attributes to describe files, and these attributes can be
used to group together files with similar properties, just as if they belonged to a single
folder, not unlike tagging, as used on social bookmarking websites such as Delicious7 or
Connotea8 is another interesting avenue for organizing documents (Hammond et al.,
2005). The big advantage of this solution is its flexibility, since the organization can now
reflect the many roles a document can play, instead of confining it to a specific place in
the file system. It also has the advantage of letting people use their own descriptors,
something which has been shown to help retrieval (Broadbent & Broadbent, 1978).
The way to navigate through files has also been the subject of new approaches, including
the use of virtual piles of documents which could be hinged, spread out and restacked
(Mander, Salomon & Wong, 1992). With the modern graphical capabilities of computers
and hand-held devices, visual metaphors such as BumpTop (Agarawala & Balakrishnan,
2006), which allows users to manipulate documents as if they were physical artifacts,
show a lot of promise. Ringel et al. (2003) tested the use of chronological landmarks
(holidays, news headlines, calendar appointments and even photographs) as retrieval aids
for finding personal documents. Lifestreams used a purely chronological method of
document access in lieu of a hierarchical structure (Fertig et al., 1996).
Another trend could be summed up as “keep everything.” This philosophy, based on the
fact that storage costs are now very low, dictates that all the documents a user produces,
receives or simply sees on the Internet should be kept in memory, in case they may prove
to be useful eventually. This approach was a core principle of Lifestreams, and has been
adopted more recently in Stuff I’ve Seen (Dumais et al., 2003), which can index all kinds
of documents: e-mails, web pages, Word files, etc. Obviously, with such a large amount
7 http://www.delicious.com/ 8 http://www.connotea.org/
13
of relatively unorganized documents, an excellent search engine is required. The real-life
application of such prototypes can now be seen in the recent Desktop Search tools
released by Google and Microsoft. Going one step further, the MyLifeBits project seeks
to digitize a person’s entire life (including paper, audio and video), and bids itself as
“fulfilling the Memex vision” (Gemmell et al., 2002).
2.2.5 Ethnography
Another area worth mentioning here is that of ethnography, and to a certain extent related
social sciences such as anthropology and sociology. Their main contribution to PIM has
been on a methodological level. In fact, naturalistic approaches derived from
ethnographic studies have become a staple of PIM methodologies (Naumer & Fisher,
2007). Observation, interviews, surveys, the think-aloud method, and other data
collection methods are frequently used, and often in a mixed-method design which can
provide triangulation, thus enhancing trustworthiness of the data.
The observation of users in their natural setting (e.g. their office) is also a key component
(Erlandson et al., 1993). Increasing importance is thus given to the contextual (Kwasnik,
1991), environmental (Neumann, 1999) and social aspects of information management
(Jacob, 2001), hopefully leading towards a more complete, holistic understanding of PIM.
The use of such methods helps give PIM researchers and software designers a better
insight into how information management is really being done, and how it could be
enhanced. Naturalistic methods are also used extensively in the user testing of PIM
prototypes and techniques (Kelly & Teevan, 2007). In fact, it could be said that
ethnography and its related disciplines are largely responsible for the widespread use of
qualitative methods in PIM (Naumer & Fisher, 2007).
14
2.2.6 Knowledge Management
Knowledge Management (KM) refers to a "systematically and organizationally specified
process for acquiring, organizing, sustaining, applying, sharing and renewing both the
tacit and explicit knowledge of employees to enhance organizational performance and
create value" (Davenport & Prusak, 1998). To reach these objectives, various tools and
strategies may be employed, such as storytelling (Kalid & Mahmood, 2010), knowledge
mapping (Jafari, Akhavan, Bourouni & Amiri, 2009) and communities of practice (Lave
& Wenger, 1991).
An important element that differentiates PIM and KM is the latter's concern with tacit
knowledge (Nonaka, 1994). Tacit knowledge can be seen as information that has been
processed and is now stored in the mind of an individual. KM seeks to capture that
knowledge and make it explicit, so that it can be communicated to others and used as
needed (Alavi & Leidner, 1999). PIM usually focuses on information as contained in
documents, even if these may be simple scraps of paper (Bernstein, Van Kleek, Karger &
Schraefel, 2008).
Furthermore, PIM focuses more on the individual's viewpoint, whereas KM is concerned
with the organization as a whole, and its ability to preserve and use existing knowledge in
a productive way (Bergeron, 2003). That being said, it is easy to imagine that if all of the
employees within an organization had excellent PIM performance, the ability of that
organization to manage (explicit) knowledge could be positively affected. So even though
they look at things from a different angle, PIM and KM ultimately share similar concerns
and objectives, and in the end appear quite complementary.
15
2.3 PIM Studies
The influence of these various disciplines has shaped PIM over the years and continues to
do so. In this section, we shall look at some of the most important PIM studies done in
the past, and highlight the main findings researchers have made in the process.
The first part of this section is devoted to the early, PIM-like studies of the 1960s and
1970s. The second part deals with the landmark workplace studies made in a paper-based
environment in the early 1980s. Finally, the third part presents some of the more recent
PIM studies, conducted in either a digital or a hybrid (paper and digital) environment.
2.3.1 Early studies
Before there was ever a mention of Personal Information Management as an area of
research, some researchers decided to study people’s informational behaviours. Very
often, the subjects of these studies were university professors or individuals involved in
research, such as scientists and engineers (Chen, 1974; Jahoda et al., 1966; Parker &
Paisley, 1966). Presumably, the focus on this type of people as subjects was due to the
perception that those were the people who had to handle the most information, a
perception that was probably not entirely unfounded. Interestingly, researchers these days
tend to study all types of people, with professors and scientists representing a much
smaller proportion of the overall samples. Perhaps that says something about the
prevalence of information in our society today.
Those early studies, despite their limitations, provided some very important insights
about their subjects’ informational behaviours, notably with regard to the channels of
communication used to acquire and disseminate information. It was found that the
informal channels or sources of information, such as conversations with colleagues, were
of great importance (Parker & Paisley, 1966). Others examined the use of journals and
books, as well as usage of the library, and how these elements fit into people’s
information landscape. Not surprisingly, the location of the material and the time required
16
to secure it were among the most important factors in information seeking behaviours
(Chen, 1974).
But beyond information acquisition, there were also interesting forays into the actual
methods used for managing all of the documents obtained or, as Case (1986) later asked:
what happens when scholars come back from the library, the laboratory or the lounge? So
perhaps more important in the context of PIM were the accounts of the personal
collections and indexing systems developed by professors and scientists. Jahoda (1966)
examined the contents of science professors’ personal indexes, which they used to keep
track of the important things they read, and that they could refer to when writing an
article or when preparing course material. Most indexes were kept in the form of note
cards or file folders, and usually consisted of article reprints, abstracts and personal notes.
Subject access was the overwhelming choice for organization, but the complexity and
elaborateness of the indexes varied widely depending on the person (see also Murphy
(1980) for an example). Soper (1976) went further by examining not only the personal
indexes, but also analyzing the personal books and journals collections that lined the
walls of professors’ offices. Among her most interesting findings were the reasons why
professors wished to amass so many documents in their offices. The first was
convenience: personal collections were just easier because they were arranged to reflect
their needs and interests. Secondly, there were no barriers between a user and his
personal collection, whereas the library materials were located farther away, and might
even be out on loan when needed. While many of Soper’s subjects expressed a certain
level of dissatisfaction with the lack of organization of their collection and with the
amount of time they had to spend on it, they still thought such collections were important.
2.3.2 The 1980s onward: the natural history of offices and the importance of context
There had been several previous studies on how people organized their documents, but
these were mostly descriptive in nature. Starting in the early 1980s, the emphasis was put
on determining why individuals classified their information in a certain way.
17
Certainly one of the most influential papers in the admittedly young history of PIM was
Malone’s (1983) article relating his investigation of office workers’ organizational
behaviours. Malone conducted interviews with workers and asked them to give him a
tour of their office. During this process, he asked questions such as “Why is that there?”
He also drew maps of the offices, depicting the position of desks, bookshelves and piles
of paper. It is worth noting that this methodological approach would become of a staple
of PIM research for years to come, and would be adapted to the digital environment. In
his investigation, Malone identified two distinct kinds of organization, neat and messy,
but noted that messiness didn’t necessarily imply disorganization. He also suggested a
possible relation between job content and neatness: subjects whose jobs were
characterised by routine, more standardised tasks (e.g. office clerks) usually had a neater
office than, for instance, research scientists, whose job was far less linear and predictable,
and who were often involved in many different projects at the same time (Landsdale,
1988). Malone then identified two major units of document organization: files (where the
elements are arranged in some systematic order) and piles (where no specific order
prevails). Finally, and perhaps most importantly, he identified two functions of
desk/office organisation: finding and reminding. Maintaining files, of course, helps us
find documents later on. But piles mostly act as reminders, reminding the user of
something that needs to be done. Putting a document in a pile is also a way to delay the
decision on where to formally classify it – perhaps the ultimate example of cognitive
economy (Rosch, 1978). Finding and reminding remain important concepts in PIM to this
day.
Another important study in the same vein was that of Cole (1982), who determined six
characteristics of people’s personal information: type, form, volume, complexity,
functions and levels of storage. This last characteristic deserves further attention.
According to Cole, there were three levels of interaction with one’s files: action
information documents, which were readily at hand and often found on piles and surfaces
close-by; personal work files, which were usually held in file cabinets and accessed
regularly; and archive storage, which was usually stored away from the office and rarely
18
accessed. A parallel can be traced here between these three stages and the document life
cycle, also referred to as the Three ages theory in archival science, in which an
organization’s documents are separated into active, semi-active (or intermediate) and
inactive documents (Rousseau & Couture, 1994).
Landsdale (1988) shed some new light on Cole and Malone’s findings, from a
psychological angle. He contended that the piles that the researchers had observed did not
reflect the users’ need for piles per se, but were in fact a compensating mechanism for the
problems of classification they faced. Therefore, he argued that the psychological
underpinnings of PIM could not be directly inferred from users’ behaviour in offices,
because that behaviour was largely adapted to overcoming the problems inherent to
organizing their information. In other words, we should not take for granted that what
people are doing now is ideal or natural, but rather view their behaviours as the best
approach they have found so far, given the tools they have in hand. In his article,
Landsdale also added a few more remarks related to the psychology of PIM, which are
worth mentioning here. First, it appeared to him that what we usually remember about a
document is not only the meaning of its content, but also contextual information such as
the document’s appearance, and what the user was doing at the time of its creation or use.
Additionally, what we remember about a particular document is often partial; therefore,
systems that require complete accuracy in retrieval will most likely fail.
Landsdale’s assertion that context was an important characteristic of documents was
confirmed by Kwasnik (1991). In her landmark study, she interviewed professors and
asked them why they had classified certain documents the way they had. She also asked
them to sort a day’s mail in her presence. She recorded and coded their answers,
obtaining a series of descriptive attributes that people used to justify their classificatory
decisions (e.g. “on the top shelf” was a location attribute, “correspondence” a form
attribute, etc.). In the end, she found that attributes related to context were the most
frequently invoked by her subjects, even before documents attributes (author, topic, form,
etc.) Interestingly, Kwasnik used her coding scheme to model the classificatory
behaviour of her subjects. Based on what they had told her, she developed classification
19
rules to handle their documents. She then used these rules to sort another day of her
participants’ mail, with impressive results. This suggests that people’s classificatory
behaviour could be modelled and predicted fairly adequately, paving the way for the
automatic sorting of documents.
Following Kwasnik’s lead, Case (1991) interviewed historians to examine the importance
of the physical characteristics of documents. He found that two physical factors, spatial
configuration of the office and document form, were often considered before topic in
determining document storage locations in the office, especially with regard to books.
Other important factors were treatment or purpose (e.g. all the biographies together; all
the textbooks on the same shelf) and even quality (“good” books being placed together).
Case also noted that individuals often start a filing system, only to abandon it after a
certain time, perhaps because it was faulty in the first place, and adapting it would be too
hard, or simply because it was just seen as too much trouble. In an earlier study, he had
also pointed out the difficulty that professors experienced in managing the notes they had
accumulated through their readings and reflections (Case, 1986).
2.3.3 PIM in the digital environment
The massive arrival of computers in the workplace inevitably led to questions regarding
the potential impacts on PIM. There was now a substantial body of knowledge on how
users managed their information in a paper-based environment, but it was necessary to
verify if the previous findings held up in the digital environment. Barreau (1995) did just
that, by conducting a modified version of Kwasnik’s 1991 study, adapted to the
computerized environment. Overall, her results were similar to Kwasnik’s
(preponderance of context as a factor in classification), although there were a few
differences, attributable in part to the influence of the software used. Nevertheless, her
subjects had kept behaviours that were more in-line with paper-based PIM. The transition
had thus been fairly easy on them, since they were able to replicate most of their usual
20
practices. On the other hand, it also meant that they were not taking advantage of the new
possibilities brought on by the computers when managing their information.
Having separately conducted similar studies, Barreau & Nardi (1995) jointly concluded
that out of the two main strategies for finding files, people overwhelmingly preferred the
location-based approach (which means that they usually know where a file is located, and
simply browse until they reach it) over the logical approach (which involves searching,
by keywords, the name of the file, etc.). They also noted that the reminding function of
documents lying on a person’s desk (Malone, 1983) was also applicable in the computer
environment, as people tended to leave files on their computer desktop for that very
reason. Finally, Barreau & Nardi distinguished three types of information: ephemeral,
working and archived. According to them, their subjects spent so much time and effort
dealing with ephemeral information that they had little concern for archival information,
which was too hard to manage, and not perceived as useful.
Barreau & Nardi’s study was criticized a year later by Fertig et al. (1996a). They argued
that the subjects’ preference for location-based access was mainly due to the fact that
tools at the time had very poor search functionalities, which explained why they weren’t
used all that much (and perhaps the situation would indeed be different today, with the
availability of desktop search applications such as Google Desktop). Furthermore,
location-based access was fine with a relatively small number of documents, but such a
strategy would not scale very well. Fertig et al. also argued that the use of the desktop for
its reminding function was due to a lack of better solutions, and that archiving
information was important, but simply too complex with today’s tools. They concluded
with the following warning: “To make general claims about information use in the
narrow scope of today’s desktop operating systems is a mistake” (Fertig et al., 1996a).
This is a good point, and one also has to be careful not to draw general conclusions based
simply on the nature of the available technology in one study setting, which might be
different in other organizations.
21
A few more studies on people’s use of file systems are worth mentioning. Carroll (1981)
examined how users named their files on their workstations. File names were limited to
eight characters in length, so Carroll was able to observe the various abbreviation
strategies and mnemonic devices used. He concluded that names were usually remarkably
structured, and obeyed a pattern defined by the user. When asked to remember what the
files contained based simply on their title, users collectively averaged a 92% success rate,
which is impressive given the length constraints they had to deal with. More recent
studies have tried to identify the types of words most often used in file names. Henderson
(2005) noted the preponderance of genre, task, topic and time, while Khoo et al. (2007)
found document type and organizational function/structure as the most likely types of
folder names.
Akin et al. (1987) studied the structure of directory spaces in a UNIX environment,
asking their subjects why they had places files in a particular directory. Relatively few
sub-directories were used, leading to a fairly flat structure, something that was also
noticed later by Barreau (1995). Several researchers have since examined worker-
developed file structures in more detail (Gonçalves & Jorge, 2003; Henderson, 2009;
Khoo et al., 2007). These structures, sometimes referred to as personal classification
schemes, can be analyzed using quantitative measures such as total number of files and
folders, depth of the structure (number of hierarchical levels), breadth of the structure
(number of folders at each level), leafiness (number of files per folder), etc. A few studies
have also investigated the effect of a structure’s characteristics on the success of file
retrieval by navigation (Bergman et al., 2010; Mas, 2007).
E-mail is also an active field of study within PIM, which is not surprising given its broad
use in the workplace, and its near-ubiquity in the work habits of some people (Bellotti &
Smith, 2000). An important early study was that of Whittaker & Sidner (1996). They
identified three main functions of e-mail: asynchronous communication, which is its main
purpose, but also task management and personal archiving/filing. They also observed
three major filing strategies for e-mail among their subjects: no filers didn’t use folders,
leaving all their e-mails in the inbox; frequent filers cleaned their inbox daily by
22
assigning e-mails that didn’t require action to the appropriate folders; spring cleaners
only periodically cleaned up their inbox, when it became too unwieldy. An advantage of
the frequent filing strategy is that the inbox can essentially be used as a to-do list. But this
strategy is perhaps more appropriate for people who receive a limited amount of e-mail,
or who are not required to be away from their desk for extended periods of time. Leaving
e-mails in the inbox has an effect similar to the reminding function observed in previous
studies (Barreau & Nardi, 1995; Malone, 1983). During the interviews, the users also
expressed strong reservations about automatic filing of e-mail (e.g. filters). Their main
concern was that they wanted to be aware of everything they received, and didn’t want to
run the risk of missing an important message. A subsequent study by Ducheneaut &
Bellotti (2001) provided additional insights into e-mail use, notably the fact that few
people used the search features to find e-mails. They relied on their filing structure, but
that structure remained quite flat, because they wanted quick access (instead of having to
navigate in a multi-level hierarchy). The authors suggested that the order of folders in the
e-mail application should be based on frequency of access, instead of the current,
standard alphabetical order. Mackenzie (2002) had also noted that users mainly relied on
their memory to find messages, with the folder titles often acting as a reminder as to
where an e-mail might be located. Neustaedter et al. (2005) observed the e-mail triage
behaviour of workers and managers, which is “the process of going through unhandled e-
mail and deciding what to do with it” (p. 1977), a task usually conducted in the morning
or when returning from a meeting, for instance. Some participants used a single-pass
approach, scanning each message only once and handling all messages in the order they
were received, while the multi-pass approach involved dealing with some types of
messages first, and then going through the list again to handle the rest of the messages.
Internet bookmarks are also of some interest to researchers since it is possible to gain
interesting insight into users’ classificatory behaviour by observing the bookmark folder
hierarchies they produce. Here, as in files or e-mail, individuals seem to vary in their
degree of sophistication, ranging from no filing at all to multi-level hierarchies (Abrams
et al., 1998). Gottlieb & Dilevko (2001) asked participants to classify sixty predetermined
web sites, and then interviewed the participants to have them explain some of their
23
classificatory decisions. They then used a coding scheme similar to Kwasnik’s (1991),
but obtained a different result, with content proving much more important than context.
This might underline a distinctive property of Internet bookmarks, which is that contrary
to file systems or e-mail folders, bookmark folders tend to be more topic-based than task-
or project-oriented. Bookmarks are more likely to be kept because they contain
information about something, information which can be useful in and of itself, regardless
of tasks or projects. In a follow-up article, Gottlieb & Dilevko (2003) tried to determine
the factors influencing the creation of folders. It turns out that the criteria guiding the
development of the hierarchy varied considerably from one user to another, which
prompted the authors to question the feasibility of a multi-user, hierarchical bookmark
system.
Boardman’s (2001a; 2001b) research also deserves mention. He observed the overlap
between the hierarchies for files, e-mail and Internet bookmarks maintained by
individuals. The overlap was greater between files and e-mail, but not as much with
Internet bookmarks. Mackenzie (2002) noted results along those lines in her study of e-
mail: some of the managers she interviewed had set up their folders system similar to the
paper filing systems they had become accustomed to in the past. They also said that they
might store e-mail messages in more than one location (e.g. paper files, hard drive or
diskettes). This is in line with Jones et al.’s (2002) study of people’s keeping behaviours
when they encounter information on the web. They found numerous strategies used by
their subjects, such as sending an e-mail to themselves, adding the URL to a personal
web page, bookmarking the page, writing the URL on paper or even doing nothing and
relying on a search engine to retrieve the page again at a later time.
Given the variety of tools and media people have to deal with, the idea of breaking down
the silos that constitute the various applications we use to manage information is certainly
appealing. But perhaps a single hierarchy for all of an individual’s documents would
result in less flexibility within each application, although Boardman (2001a) suggests that
the use of filtering might help adapt this master hierarchy to the various applications.
24
Does the prevalence of computers mean that PIM can now only concern itself with the
digital environment? Not necessarily. There is still a significant amount of paper lying
around in people’s offices. For one thing, people have a strong tendency to print
electronic documents. Furthermore, Jones & Teevan (1997) reported that the use of pen-
and-paper devices such as calendars, to-do lists and personal organisers/diaries was still
much higher than the use of PDAs and electronic organisers. This was confirmed by
Bellotti & Smith (2001). As they noted, new PIM tools would have to find better ways to
accommodate paper, or else find a way to adequately replace paper, which they describe
as “a high resolution, intuitive, lightweight, cheap and flexible medium” (p. 232). Of
course, the situation might evolve differently in the coming years, with newer and better
tools, and with a generation of native computer, video game and cell phone users, who
may be much more comfortable than their parents in an all-digital environment. A further
concern about digital PIM tools is that, as Neumann (1999) notes, there is very little
research or discussion on how spatial elements are lost in the computer environment; that
is, we don’t quite know the implications of actually replacing physical spaces with virtual
spaces.
2.4 Context of the present study
As we have seen so far in this chapter, managing information is not a new challenge, but
it's one that appears to be increasingly important, notably in the workplace. This state of
affairs confirms the importance of PIM, which can help better understand users'
behaviours, and in turn inform the development of appropriate tools to accomplish their
tasks.
Drawing from several disciplines such as cognitive psychology, human-computer
interaction and ethnography, PIM has evolved through the years, from its first studies in a
paper-based environment, to the appearance of the first micro-computers in the
workplace, up until today's largely digital information landscape. Along the way, various
types of information management applications were studied, such as operating systems,
personal digital assistants and e-mail.
25
Overall, two main types of PIM research appear to be especially prominent: studies of
users' PIM behaviour and studies testing the efficiency of new PIM prototypes. In a way,
these two types of studies are complementary, since the results of one can inspire the
other, and vice versa. Such studies have used both qualitative (e.g. interviews or
observation) and quantitative measures (e.g. retrieval efficiency).
So even though PIM is a relatively young field of study, a sizable body of work has been
accumulated. In our view however, three aspects in particular deserve more attention than
they have received so far in the literature.
First, in recent years, a strong focus has been put on quantitative measures, such as the
breadth or depth of the structure, the number of files per folder, etc. (Gonçalves & Jorge,
2003; Henderson, 2009; Khoo et al., 2007). While such measures are interesting, it seems
obvious that there is still room for more qualitative research in order to better explain the
reasons behind these observations. Folder structures are built by human beings making
decisions in very specific conditions. Qualitative methods appear especially suited to
understanding such phenomena.
Secondly, it is possible to view a person's document structure as the result of a multitude
of individual classificatory decisions (putting a document here, creating a new folder
there, etc.). By examining - and asking about - specific instances of document
management decisions, we are likely to gain insight into the mechanisms by which folder
structures are built and maintained. Therefore, perhaps we need to go beyond a
superficial, general view of a person's structure and take a more micro approach.
Thirdly, very few of the studies mentioned here concerned workers in support staff
positions (Malone, 1983; Mas, 2007). The lack of available data on such an important
part of today's workforce confirms the need for such research.
26
Our objective is thus to try and address these aspects in our study, both in our choice of
target population and our choice of methodological approach.
2.5 Information, documents and files: terminological considerations
In the literature reviewed above, there appears to be a great deal of variation in
terminology, especially with regard to the terms information, document and file. In fact,
these words often seem to be used as equivalents or quasi-synonyms. For example,
authors sometimes refer to managing information (Czerwinski et al., 2006; Pratt, et al.,
2006), managing documents (Bondarenko & Janssen, 2005; Henderson, 2004), or
managing files (Indratmo & Vassileva, 2006) to describe very similar realities.
On this issue, Pédauque (2003, p. 2) states that "[a] very large number of research papers
use different vocabularies, rigorously defined in some cases but often subject to different
interpretations, to designate comparable objects." And this issue is by no means the
province of PIM alone. Tyrväinen & Päivärinta (1999, p. 1) contend that "the
contemporary literature in information systems (IS) research carries a rich collection of
implicit connotations for a 'document', hence lacking a solid theoretical foundation and
commonly accepted terminology."
Part of the problem may be that these are words we use daily as part of our vocabulary,
but also words that refer to somewhat intuitive concepts. At the very least, it is fair to say
that information, documents and files are closely related notions, even though they are
not exactly equivalent. In this section, we will attempt to define those terms, and to
clarify the relationship between them. Our aim is not to provide a definitive solution to
the current terminological conundrums, but merely to establish what these terms mean in
the context of our research.
27
2.5.1 Information
Information has been defined in various ways in the past, depending in large part on the
approach or the discipline concerned. Statements such as "information transfer =
communication" or "means by which uncertainty is reduced" have obtained a wide degree
of acceptance (Hoffmann, 1980, p. 291). In recent years, the idea that information carries
meaning has become more prominent (Floridi, 2005). Information could thus be defined
as "data that has been processed into a form that is meaningful to the recipient" (Davis &
Olson, 1985, p. 200). The notion of meaning also helps separate information from data in
the famous Data-Information-Knowledge-Wisdom paradigm (Hey, 2004). In our study,
we will retain this notion of meaning as a fundamental property of information.
2.5.2 Documents
The word document has also been given many definitions through the years, perhaps even
more so since the advent of electronic documents. Levien (1989) defined a document as a
unit of "recorded information structured for human consumption", which Sprague (1995)
revised as "an information set pertaining to a topic, represented by a variety of symbols,
stored and handled as a unit". The main ideas here are those of structure and unity, and
also of an object with well-defined boundaries. Another widespread view is that of the
document as a "container of information" (Ferraioli, 2005, p. 77).
These definitions are not as well identified in the digital environment. Karger & Jones
(2006) contend that it has become harder to distinguish what constitutes a document, due
to the fragmented nature of electronic documents, which are often spread among several
devices or software applications. Schamber (1996, p. 669) raises several questions in that
vein: "Do linked portions of several documents constitute a single document? Is a thread
of e-mail messages a single document? Who owns the intellectual content of a set of
hyperlinked documents? Of a collaborative document?" In short, boundaries are
becoming fuzzier. The idea that a document usually has a well-defined structure is also
28
debatable. Do a few notes jotted in Notepad constitute a structured document? What
about the ability to cut and paste text at will, and thus recombine elements from different
sources, altering the original document(s) and creating new ones? Those possibilities
seem to undermine any definition of a document as a rigid, inflexible object. In this
context, the document appears as a more or less structured, more or less bounded,
container of information. This definition might be a bit vague, but it is in line with our
somewhat fractured digital landscape, and is probably sufficient for the needs of our
study.
2.5.3 Files
A further distinction can be made between documents and files. The word file here refers
to computer files, which is what workers actually manipulate on their computer
workstations (not to be confused with the files in Malone’s files/piles dichotomy, which
refer to a collection of documents often contained in file cabinet). Computer files can be
opened, saved in directories, moved from one directory to another, downloaded, sent via
e-mail, etc. The distinction between documents and files is underlined by Dourish et al.
(1999): "Documents [...] are not files. Documents are organizational entities with social
meaning and practical consequences, while computer files are computational artifacts
whose behaviors are defined technically" (p. 134). In other words, the file is the actual,
physical incarnation of a document.
A typical office workstation mainly contains two types of files: some are files that
embody documents, as described above, while some are system files. The purpose of
system files is not to convey meaning; they are simply there to ensure the working order
of the computer. In a Windows environment, they would have extensions such
as .exe, .dll, .bin, etc. System files can theoretically be managed by workstation users, in
the sense that they can be moved around or deleted, but they are not necessarily
meaningful to the average worker, in that they don't carry usable information in the
context of work.
29
In this study, we shall focus on the evaluation and manipulation of files that are
documents, such as Word documents, e-mails, still images, audio and video files, etc. It is
worth noting that if we exclude non-document files from consideration, the distinction
between documents and files in an electronic environment becomes very subtle, to the
point where those two terms can be used almost interchangeably. The only remaining
distinction is on a semantic, rather than practical, level. Figure 2.1 below is a
representation of the distinction we make between information, document and file in the
context of our study.
Information: the meaning /
message
Document: the form /
structure
File: the object /
artifact
25 new clients; $1 million
in additional revenue, etc.Progress
Report
Jan. 2006ProgRep
Jan06.doc
Figure 2.1: Information, document, file
2.6 Conceptual framework
Many of the studies examined in this chapter follow a similar logic, with the basic
assumption that through observation of a worker’s documents, and through other means
such as one-on-one interviews, it is possible to eventually understand what factors have
influenced the development of a specific document structure. This is essentially what we
are trying to accomplish in this study.
30
In the present section, we present a basic conceptual framework for our study, based upon
some of the key findings in the literature. This framework will serve as an important
guide in the development of our initial interview questions.
2.6.1 Context of document organization
Barbara Kwasnik was among the first to highlight the fact that contextual factors can
influence the way a document is classified by a worker. She identified several such
factors in her 1991 study entitled "The importance of factors that are not document
attributes in the organization of personal documents".
Kwasnik found that workers didn't necessarily follow what could be considered a
traditional classification scheme (e.g. by topic, author or title). Such considerations,
which she called "document attributes", could often be superseded by other aspects such
as the intended use of a document or its perceived value in the context of a specific
project.
This has important implications for our study. It means that in order to be able to
understand why certain documents have been classified a certain way, it is necessary to
investigate the phenomenon on a broader level. It thus becomes crucial to go beyond the
content of the documents and try to understand the context in which these documents are
used or created. As such, a large portion of our initial interview questions should focus on
contextual aspects. It will be important to understand the environment in which our
participants work, and what their role is within that environment. This could include their
place within the organizational hierarchy, or the people they interact with on a daily basis.
One specific aspect of the workers’ context mentioned by Mas (2007) is the presence of
an institutional classification scheme mandated by an organization. Obviously, this could
potentially influence the way workers handle their documents.
31
2.6.2 Job content
Malone (1983) established that the nature of a worker’s position, and the tasks they have
to perform, can have a significant impact on document organization. As we mentioned
earlier, workers with more standardized, routine tasks, generally have a more neatly
organized office. At the other end of the spectrum, people whose tasks are more varied
and unpredictable tend to have a messier, less structured organization.
Beyond this relatively simple dichotomy, every worker’s job is to a certain extent unique,
which could translate into differences in document organization. In the workplace,
documents are normally created and used in order to accomplish some type of result: to
communicate information, to record data or facts, to support organizational processes, etc.
(Sprague, 1995 ; Hertzum, 1999). Documents are thus intimately linked to the tasks they
are meant to support. Therefore, when studying a worker’s documents, it is also
necessary to pay attention to the specific nature of their tasks and responsibilities.
2.6.3 Document attributes
Although Kwasnik (1991) highlighted the importance of contextual factors in
determining document organization, document attributes (such as their topic or form)
were still considered very significant by her participants, and eventually had an incidence
in document organization. These attributes should therefore not be ignored.
In the same study, Kwasnik also talked about the notion of value. Some documents were
deemed especially important or valuable, which could translate into special treatment or
consideration. There has been further evidence in the literature that the perceived value of
a document may affect the way it is handled (Bondarenko & Janssen, 2005; Paré, 2007;
Whittaker & Hirschberg, 2001).
32
In short, the properties of individual documents must also be taken into account when
trying to understand how workers organize documents. It is thus necessary to both
observe the general context and consider the decisions workers make at the document
level.
2.6.4 Individual preferences and abilities
According to Gwizdka and Chignell (2007, p. 213), “information organization strategies
appear to be related to the user’s retrieval preferences and style”. This translates into
dichotomies such as filers vs. pilers (Malone, 1983) or frequent filers vs. spring cleaners
(Whittaker & Sidner, 1996), which have not really been explained satisfactorily, but still
represent a documented reality. It seems that the best we can do at this point is to label
these as “individual preferences” with respect to PIM.
Gwizdka and Chignell (2007) also allude to the potential impact of cognitive abilities in
PIM, most notably memory and spatial abilities. One would thus expect that people with
different capabilities in that regard could develop different PIM strategies and behaviours.
Finally, Mas (2007), citing Chapman (1999), proposes that experience using computers is
a potential factor affecting the development of personal classification schemes.
2.6.5 Towards a conceptual model
Figure 2.2 below illustrates how the factors mentioned in this section form the basic
conceptual framework for our study.
33
Context (work environment)
Individual documents: - Attributes - Value
Individual preferences and abilities
Job content
(tasks)
Document organization
Figure 2.2: Conceptual framework, as derived from the literature.
This conceptual framework, while useful, should neither be seen as a formal hypothesis,
nor as a rigid frame for our research. This is especially important given our chosen
methodological approach, which is grounded theory. Therefore, the conceptual
framework presented herein should be considered more akin to Glaser’s (1978)
sensitizing concepts. According to Charmaz (2003, 259), sensitizing concepts are
“background ideas that inform the overall research problem”. Such concepts can thus
serve as starting points, but shouldn’t be seen as limiting in terms of the scope of the
research, or the specific topics that should or should not be addressed in interviews
(Bowen, 2006). They can however inform the initial development of the interview guide.
34
Chapter 3: Methodology
This chapter presents a detailed account of the methods used in the course of this research.
A rationale for the use of grounded theory is first given, after which the research setting
and sample are presented. The data collection process is then explained, followed by a
brief discussion of validity concerns.
3.1 Grounded Theory
Given its objectives and its sampling and data collection methods, the present study
clearly positions itself within the qualitative paradigm. The particular methodological
approach chosen for this study is that of grounded theory (GT). Grounded theory is not a
qualitative methodology per se, as it can also be used to make sense of quantitative data.
Indeed, the creators of grounded theory considered it as a general methodology (Glaser &
Strauss, 1967). Nevertheless, grounded theory is most often used to analyze qualitative
data obtained through interviews, observations and document analysis. It also uses a
qualitative approach to analysis. Grounded theory has been described as "a
comprehensive and systematic framework for inductively building theory" (Punch, 2000,
103). Developed by Glaser and Strauss (1967), it includes a set of methods and guidelines
to help collect and analyze data in order to understand a given situation, and eventually
derive appropriate theoretical propositions. Its emphasis is thus on generating new theory
from the data obtained, instead of trying to verify an existing theory or hypotheses.
According to Stern (1995), "the strongest case for the use of grounded theory is in
investigations of relatively uncharted water, or to gain a fresh perspective in a familiar
situation", which adequately describes the context of our study.
The origins of grounded theory can be traced back to the theoretical inclinations of its
two creators. Glaser was influenced by the positivist school of thought at Columbia
University. He was trained in quantitative methodology and qualitative mathematics,
which strives to express quantitative data in qualitative terms (Urquhart & Fernandez,
35
2006), and through grounded theory, strived to develop rigorous methods that could be
used within a qualitative paradigm (Charmaz, 2006). Strauss, on the other hand, came
from the Chicago school of pragmatism and symbolic interactionism, which assumes that
“society, reality and self are constructed through interaction and thus rely on language
and communication” (Charmaz, 2006, p. 7). The combination of these two approaches
eventually informed the creation of grounded theory.
In the years following the publication of their seminal 1967 book, Glaser and Strauss
adopted slightly different views of grounded theory. It could be said that Glaser stayed
closer to the original principles, while Strauss, along with Corbin, introduced a new set of
techniques devised to help analyze data, moving away from the basic constant
comparison method originally advocated (Strauss & Corbin, 1990). This was seen by
Glaser as attempts at forcing the data into preconceived categories (Charmaz, 2006).
The differences between the two original authors’ approaches have yet to be reconciled.
In fact, new approaches to (and variations of) grounded theory have surfaced since and
continue to be proposed. Perhaps the most prominent among those is that of Charmaz.
Inspired by post-modernism and post-structuralism, her approach is a constructivist one,
in opposition to the more positivist leanings of both Strauss and Glaser (Lomborg &
Kirkevold, 2003). Charmaz assumes that theories are not discovered, but rather
constructed “through our past and present involvements and interactions with people,
perspectives, and research practices” (Charmaz, 2006, p. 10), a view we tend to share.
3.2 Data Collection and analysis
Data collection for our study consisted mainly of interviews and observation. Grounded
theory does encourage the use of more than one data collection method when possible. In
fact, according to its creators, the more data can be obtained, and from the most sources,
the better (Glaser, 1998). Any new element of data can be integrated into the research,
coded and compared to previous data in order to help generate new theory.
36
The main method of data collection used here was semi-structured interviews, in which
there were not necessarily fixed, precisely ordered questions, but rather a few questions
that were used as a starting point, with probes and follow-up questions added along the
way, as appropriate. The idea was to give the participants enough freedom to express
themselves, while still making sure that all the topics of interest were covered. The
interview guide appears in Appendix 2. The follow-up questions ensured that the
interviewer fully understood what the participants were saying, in case something wasn’t
entirely clear initially. This ongoing clarification and confirmation of what was being
expressed was a significant help in improving the trustworthiness of our results.
The planned duration of each interview was about 45 minutes to an hour, the same as the
interviews conducted during a previous pilot study 9 . In general this was respected,
although durations actually ranged from 30 minutes to 1:45. The earlier interviews tended
to be longer, since we wanted to make sure every topic of interest was covered. Later
interviews were shorter, with a few exceptions, as the focus of the interviews was
sometimes narrower. In some cases, only very specific aspects were tackled, in order to
confirm or contrast with previous interviews. Some of the topics had become
theoretically saturated and were simply dealt with in passing. We should mention that in
all cases the participants were very generous with their time, and in several cases
interviews went longer than the planned length, even though participants were told that
the planned interview duration was over. During the interviews, some handwritten notes
were taken, although in limited amount. Intensive note-taking tends to distract from what
is being said, and can cut the flow of the conversation, for example when a participant
slows down or stops talking to let the interviewer catch up (Ary, Jacobs, Razavieh &
Sorensen, 2009; Mariampolski, 2001). The interviews were therefore digitally-recorded
and later transcribed.
A few days before each interview, participants were invited to fill in a short online survey,
which was developed on the SurveyMonkey platform10. The main purpose of this survey
9 More details on this pilot study are available in section 3.2.1. 10 http://www.surveymonkey.com
37
was to help the researcher gather relevant contextual information about each participant,
in order to better prepare the interviews. Furthermore, because of the limited time
available for each interview, we felt it was better to gather this type of information in
advance, instead of taking too much interview time for demographic questions. Although
it can be a good idea to ask simple, factual questions at the beginning of an interview in
order to make participants at ease, we wanted to focus the bulk of the interviews on
participants’ PIM behaviour. Some simple questions were asked at the beginning of
interviews, mainly in order to get a better understanding of the nature of the workers’
position and responsibilities, and to establish rapport. The survey we used was actually an
adapted version of the one used by Sabine Mas in the study she conducted with office
support staff in a very similar setting (Mas, 2007)11. Our version of the survey, which
was used in this study, is presented in Appendix 3. The main results of the survey are
presented in section 3.3.
Visual observation was also used in the course of the interviews. The participants were
interviewed in their office, which added an important visual component to the verbal
description of their document organization behaviours. Participants illustrated their
answers by demonstrating how they had organized their files, how their predecessors
(whose files they had kept) used to do it, etc. Other visual cues such as general office set-
up and level of clutter provided the researcher with additional insights, and often
prompted additional questions. The method of interviewing participants in a natural
setting - their office - is quite common among PIM and HCI (Human-Computer
Interaction) researchers, and has proven successful in the past (Ducheneaut & Bellotti,
2001; Malone, 1983). Screenshots of the contents of key directories or other interesting
elements from their computer workstations were made, in order to help in the analysis
phase. Because their contents may help reveal the identity of the participants, these
screenshots will not be shown here.
The data obtained was analyzed as follows. The first phase consisted of reading the
interview transcripts and written field notes and systematically coding their contents 11 We wish to thank Dr. Mas once again for allowing the use of her survey questions.
38
using the NVivo 7 software from QSR12. In grounded theory, unlike other methodologies,
there is no predetermined list of codes. Codes are instead created during analysis, as
needed to represent the reality observed. In vivo codes can also be used. This process is
sometimes referred to as open or substantive coding (Heath & Cowley, 2004). Table 3.1
provides a few examples from this first phase of coding. In total, 181 different codes
were used a total of 954 times.
Code used Example
Lack of time That file right here, I should have put it in the "Accueil" folder, but I just didn't have time to do so.13
Thinking about your successor
But eventually, if I had to leave, I would surely review my system to make sure that the next person can find her way around.14
Printing
The documents he needs for his meetings, that, I'll print.15
Arriving in a New Position
I haven't been here since the beginning, so I don't know the full story behind this project.16
Table 3.1: Coding examples (initial phase)
The second phase, sometimes called axial coding, involves the creation of categories and
sub-categories from these codes, for example by merging or grouping existing codes
together. In the final phase, the researcher builds a framework from the categories
obtained earlier. This method ensures that the emerging theory is as representative of the
collected data as possible.
12 http://www.qsrinternational.com 13 “Ce document-là ici, j'aurais pu le mettre dans le dossier Accueil, mais j'ai juste pas eu le temps.”
[Béatrice] 14 “Mais c'est sûr qu'éventuellement, si j'avais à quitter, je reverrais mon système pour faire en sorte que la
personne qui arrive puisse se retrouver.” [Béatrice] 15 “Les documents dont il a besoin pour ses réunions, ça je lui imprime.” [Jeanne] 16 “Parce que je ne suis pas là depuis le début, alors je ne connais pas toute l'histoire de ce dossier-là. ”
[Lise]
39
As we mentioned earlier, grounded theory stipulates that data collection and analysis are
a simultaneous process, where new data is almost immediately analyzed, integrated and
compared with the mass of data already analyzed. By using the constant comparative
method, concepts are constantly merged, renamed and modified. Furthermore, at any
time during the data collection process, the writing of memos is encouraged in order to
make note of emerging concepts and to reflect on their meaning at that stage of the
research. Several such memos were written in the course of our research.
Because our study took place in a French-language university, the interviews were
conducted in French, which is the researcher’s and the participants’ native language. The
coding, however, was done in English. Great care was of course taken to ensure that the
resulting concepts accurately represented participants’ ideas and opinions. In the
following chapters, quotes from participants are presented in the text translated in English,
but the original quotes are available as footnotes. This was done to ensure a more uniform
reading experience, while still allowing those who might be interested to read the original
version.
3.2.1 Pilot study
In 2006, we conducted a small-scale pilot study to investigate the PIM behaviour of
workers in an academic setting. The purpose of this pilot was to determine whether the
envisioned methodology - the combination of observation and interviews – would be
adequate for our purposes. Although such methods had been employed in previous PIM
studies, our intent was to validate their use in the context of our study. It was also a way
for the researcher to become more comfortable with the conduct of such interviews.
In the end, 5 participants were interviewed: two librarians, two managers and one office
support staff member. It is important to note that when the pilot was conducted, our
research focus was not yet solely on office support staff, but rather on academic workers
in general. The pilot was, however, set in the same organization as the present study.
40
One significant benefit of the pilot was that it helped us enhance the interview guide for
the present study. More specifically, it was instrumental in determining the optimal
length of the interviews as well as the appropriate sequence of questions, and brought to
our attention several new topics of investigation. It also helped with more technical
aspects such as the ideal position of the digital recorder during the interview. Results
from this pilot study were later presented in Paré (2006).
The pilot study also deserves mention here because some of the data gathered during the
pilot were reused in the present study. Indeed, in grounded theory, it is always possible to
use existing documents to add to the mass of relevant data within a study (Charmaz,
2006). Since one of the participants in the pilot was an office support staff worker who fit
all of this study’s requirements, we decided to integrate her interview transcript among
the data to be analyzed. She is listed here as Francine, participant #6.
3.2.2 Ethics considerations
Both the pilot study and the main study were approved by the Ethics Review Board of
McGill University’s Faculty of Education, in accordance with the requirements of the
McGill University Policy on the Ethical Conduct of Research Involving Human
Subjects17, and the Tri-City Council Policy Statement on the Ethical Conduct of Research
Involving Human Subjects18.
17 http://www.mcgill.ca/files/secretariat/Ethical-Conduct-Research-Human-Policy.pdf 18 http://www.pre.ethics.gc.ca/eng/policy-politique/initiatives/tcps2-eptc2/Default/
41
3.3 Setting and sample
For this study, we chose to look more closely into the personal information management
behaviours of support staff in an academic environment, more specifically office workers
in a university setting. The academic setting is an interesting one in which to conduct
such a research, since universities are usually complex, document-heavy environments,
which thus provided a potentially rich environment for our study. In that sense, it was
also our hope that this study could have a positive impact on the way electronic
documents are managed in universities. The chosen setting for our study was a large
French-language university located in Canada.
In our target setting, participants occupying the following positions were recruited:
Secretarial Agent, Administration Technician, Office Work Technician, Student Files
Technician and Secretarial Coordinator. These are relatively similar positions in terms of
their work environment, the kind of work they accomplish, and their relative place within
the organization. Here is a translation of the official job descriptions for these positions,
as obtained from the University’s Human Resources Department:
SA - Secretarial Agent (Agent(e) de secrétariat, abbreviated as ‘AS’)
Ensures the support and follow-up of administrative processes in his/her unit.
Welcomes, informs and helps users. Types and lays out documents and forms
(paper or electronic); verifies the quality of the French writing and the
accuracy of data. Files, photocopies and assembles documents.
AT - Administration Technician (Technicien(ne) en administration,
abbreviated as ‘TA’)
Collaborates in the elaboration, application and control of administrative
standards and procedures. Ensures the realization of projects and the smooth
functioning of administrative, financial and material operations he/she is
assigned.
42
OWT - Office Work Technician (Technicien(ne) en coordination de travail de
bureau, abbreviated as ‘TCTB’)
Ensures the support and follow-up of current operations and the smooth
functioning of the office, either in a faculty or an administrative service.
Maintains his/her superiors' agenda, follows up on correspondence (mail, e-
mail and telephone). Participates in planning and organizing the activities of
the faculty or service, such as meetings, conferences, committees and travels.
Prepares and revises the necessary documents for these various activities.
Keeps accounting and staff files up to date.
SFT - Student Files Technician (Technicien(ne) en gestion des dossiers
étudiants, abbreviated as ‘TGDE’)
Ensures the management of student files according to the program structures
and the current academic rules. Informs and counsels students during the
entire course of their studies, from admission to graduation, at every level.
Updates information in the University Portal. Controls and processes choices
and requests made by students.
SC - Secretarial Coordinator (Coordonatrice de secretariat, abbreviated as ‘CS’)
We were unable to obtain an official job description for this position, but as
far as we can gather, the tasks involved are very similar to those of an office
work technician (OWT). However, the Secretarial Coordinator can have
additional managerial duties, such as overseeing and coordinating the work of
other OWT or SA workers within a department. Because of this, Secretarial
Coordinators are not members of the office workers’ union, as opposed to the
four positions presented above.
Grounded theory advocates for the use of theoretical sampling, a form of purposive
sampling in which the goal is "to help refine and develop the theoretical concepts that are
emerging from the analysis" (Weed, 2009, p. 505). It is important to understand that in a
43
grounded theory study, data collection and analysis are done more or less simultaneously,
in a sort of iterative process. Coding, identification of concepts and even theory
development usually begin very early on. In fact, data analysis may start as soon as some
data is collected (e.g. a single interview in our case). Subsequent interview participants
can then be chosen for theoretical purposes. For example, the researcher may wish to
interview someone in a similar position in order to compare and confirm previous
findings, or may want to find someone with a totally different profile in order to see if the
results are indeed different.
Theoretical sampling is thus instrumental in allowing the researcher to employ the so-
called constant comparative method throughout the analysis. Constant comparison
between emerging codes and concepts is encouraged in order to refine the emerging
theory and ensure its comprehensiveness. In fact, as opposed to other methodologies,
where one may wish to have some level of uniformity among participants, the grounded
theorist will most likely want to increase the diversity of the sample in order to get a
more complete, all-encompassing view of the studied phenomenon.
According to Strauss & Corbin (1990), "[s]ampling, analysis and theory building are to
be repeated until the goal of explaining the phenomenon at hand has been reached”. The
stopping point occurs when the data becomes saturated, that is when no new concepts
emerge from the data, or when new data does not change the emerging theory anymore.
Because of this, it is usually hard to determine in advance the precise number of
participants one needs to reach. A survey of grounded theory studies by Thomson (2004)
showed a great variety in sample sizes, although theoretical saturation generally seemed
to occur somewhere between 10 and 30 participants. Therefore, our initial target was to
reach around 10 to 15 participants, with additional participants added as needed. Beyond
a specific number of participants, our main concern was to obtain theoretical saturation.
In the end, 15 interviews were conducted before we were confident that a saturation point
had been reached.
44
The recruitment process went as follows. Participants were contacted by e-mail, and
asked if they were willing to participate in our study. An example of such an e-mail is
presented in Appendix 1. The first few participants were recruited within the same
department. Those first participants allowed us to gather our first mass of data, which
formed a basis for later comparisons. Later participants came from various other
departments and were usually chosen because of specific characteristics we wanted to
explore more fully, such as department or service affiliation, age or years of experience,
years in the current position, workers using the University’s institutional classification
scheme, people in interim positions, etc. Sometimes participants suggested people they
knew, for example collaborators and current or former colleagues, as potential
participants.
Because of logistical constraints, the interviews were conducted over the course of
several months, and thus were scattered across the university time cycle (Fall, Winter and
Summer terms; beginning, middle and end of the term, etc.). Although it was useful to
keep this contextual element in mind when conducting the interviews, the fact that we
couldn’t control this variable was not a major issue. One reason is that we were not
looking to compare identical situations in our study: it was expected that every
participant would work in a relatively unique context anyway, of which the time period
would be but one element. However, we certainly had to consider this temporal element
as a possible explanation for some of the behaviours encountered.
Table 3.2 below presents an aggregated list of the participants in our research. It should
be noted that all the participants interviewed in this study were women. This is not so
much a choice as a consequence of the makeup of our target population, in which women
represent an overwhelming majority. Indeed, data obtained from the university’s Human
Resources Department reveals that men accounted for a bit less than 3% of the total
number of employees in the targeted positions at the time of our study. For example,
none of the university’s 82 secretarial agents were men. Efforts were made to include at
least a few male participants in the study, without success. Nevertheless, we believe that
45
46
our sample remains fairly representative of the population as a whole, even though the
uniformity in gender does deserve a mention here.
# Participant name
Age Years in the
university
Position Years in this
position
Use of official scheme
Training Self-assessed computer literacy
Self-assessment of
document organization
Level of difficulty in organizing documents
Interview date
1 Anne 30-39 6-10 OWT 1-5 No Secretarial College
Expert Very well Very easy 23-Jul-08
2 Beatrice 40-49 6-10 SC 6-10 No Cegep + Archives
Intermediate Well Moderately easy
24-Jul-08
3 Carole 40-49 6-10 SA 1-5 No Archives Intermediate Well Moderately easy
29-Jul-08
4 Diane 40-49 16+ OWT 0-1 No Archives Expert Moderately well
Easy 12-Aug-08
5 Édith 30-39 1-5 OWT 1-5 No Cegep + Archives
Expert well Easy 4-Aug-08
6 Francine19 50-59 16+ OWT 6-10 No - - - - 8-Apr-06
7 Ginette 30-39 6-10 SFT 1-5 No Archives Expert Very well Easy 19-Nov-08
8 Hortense 50-59 16+ OWT 1-5 No Archives Intermediate Well Easy 19-Nov-08
9 Ingrid 20-29 1-5 AT 0-1 No Cegep + Archives
Expert Well Easy 29-Nov-08
10 Jeanne 20-29 1-5 OWT 1-5 No No Expert Well Easy 10-Dec-08
11 Kim 40-49 6-10 OWT 1-5 Yes Archives Intermediate Moderately well
Easy 17-Dec-08
12 Lise 20-29 1-5 OWT 0-1 No No Intermediate Well Easy 12-Aug-09
13 Manon 50-59 16+ AT 1-5 Yes Archives Expert Moderately well
Moderately easy
26-Aug-09
14 Nicole 40-49 6-10 SA 6-10 No No Intermediate Well Easy 18-Sep-09
15 Olivia 20-29 0-1 OWT 0-1 Yes Cegep Intermediate Moderately well
Moderately easy
23-Sep-09
Table 3.2: Participants in this study.
19 Pilot study participant.
47
In the following pages, each column in this table is explained, with additional details
given as necessary. In relevant cases, a graphical representation of the data is also
provided.
# - Each participant was given a number in order to facilitate identification. The number
corresponds to the order in which data for that interview were analyzed.
Participant name - Although the present research doesn’t deal with very delicate subject
matters, it was still important to make sure that the participants couldn’t be readily
identified. In order to ensure participants’ anonymity, their names were replaced by
pseudonyms, which will be used in the remainder of this text. Likewise, participants’
specific department or unit is not indicated herein, as in some cases this could have
allowed for the identification of the individuals in question.
Age - The estimated age of each participant. Figure 3.1 presents a graphical
representation of the age distribution. The graph shows that we were able to interview
participants from various age groups, which was one of our initial objectives.
0%
5%
10%
15%
20%
25%
30%
35%
20-29 30-39 40-49 50-59
Estimated Age
Perc
ent
Figure 3.1: Frequency distribution of age
48
Years in the University – The number of years the employee had worked at the university
where we conducted this study. Figure 3.2 presents a graphical representation of the years
of employment at the university.
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
0 to 1 1 to 5 6 to 10 11 to 15 16+
Perc
ent
Years in the University
Figure 3.2: Frequency distribution of the years of service in the University
Position – The position held by the employee. These abbreviations were explained at the
beginning of section 3.3.
Years in this position – The number of years the employee has been working in his/her
current position. Figure 3.3 presents a graphical representation of the years in the current
position.
49
0%
10%
20%
30%
40%
50%
60%
0-1 1-5 6-10
Years in position
Per
cent
Figure 3.3: Frequency distribution of the years of service in the current position
Use of official scheme – Indicates whether the employee uses the university’s official
document classification scheme in his/her current position. This system was devised and
is maintained by the University’s Archives Division. Three of the participants used the
system at the time of our study.
Training – Indicates the type of training or education the employee has received with
regard to document organization. It does not consider other types of training or diplomas.
A ‘+’ sign indicates that the participant has received more than one type of training. Here
is a brief explanation of the types of training listed in the table:
Archives - Basic document management training offered by the University's
Archives Division. It usually lasts one day, and is especially targeted to office
support staff, although it is not mandatory. Additional training is also available on
specific aspects, such as the University's institutional classification scheme.
50
Cegep - CEGEP is an acronym which stands for "Collège d'enseignement général
et professionnel". This type of college, which is unique to the Quebec education
system, follows high school and precedes university. It usually offers two main
paths - a two-year pre-university program or a three year vocational program - but
several one-year specialty programs also exist. Programs pertaining to office
automation and office work coordination are a staple of such institutions.
Secretarial College - A specialized, high school-level institution which trains
students in secretarial duties.
No - No specialized training in document management.
Figure 3.4 shows a graphical representation of the training profiles of our participants.
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
No training Archives Secretarial College
Cegep Cegep + Archives
Perc
ent
Type of training
Figure 3.4: Frequency distribution of formal training.
Self-assessed computer literacy - In our pre-interview survey (question C3), participants
were asked to evaluate their own level of computer literacy. The possible answers were:
51
Novice - I try to use computer applications, but I frequently need help.
Intermediate - I can execute simple tasks on a limited number of software
applications.
Expert - I am very competent in using a great number of software applications.
Figure 3.5 shows a graphical representation of the participants’ self-reported computer
literacy.
0%
10%
20%
30%
40%
50%
60%
Novice Intermediate Expert
Computer literacy
Perc
ent
Figure 3.5: Frequency distribution of self-assessed computer literacy.
Self-assessment of document organization - In our pre-interview survey (question H1),
we asked participants to evaluate how well-organized their documents were. The possible
answers were: Very disorganized, Disorganized, Moderately well-organized, Well-
organized, and Very well-organized.
52
Figure 3.6 shows a graphical representation of the participants’ self-assessment of their
document organization.
0%
10%
20%
30%
40%
50%
60%
Verydisorganized
Disorganized Moderately well-organized
Well-organized Very well-organized
Organization of documents
Perc
ent
Figure 3.6: Frequency distribution of the participants’ self-assessment of their
document organization.
Level of difficulty in organizing documents - Question H2 of the pre-interview survey
asked participants to evaluate how difficult they considered the task of organizing
documents. Possible responses were: Very difficult, Difficult, Moderately easy, Easy, and
Very easy.
Figure 3.7 shows a graphical representation of the participants’ level of difficulty in
organizing documents.
53
0%
10%
20%
30%
40%
50%
60%
70%
Very difficult Difficult Moderately easy Easy Very easy
Difficulty level
Per
cent
Figure 3.7: Frequency distribution of the participants’ level of difficulty in
organizing documents.
Interview date- Indicates the date when the interview with that participant was conducted.
The pre-interview questionnaire (question B7) also helped us gather information about
the nature of each participant’s tasks. This is important because it helps situate the
context in which employees work on a daily basis. Table 3.3 below gives a full list of the
tasks encountered, along with the number of participants who accomplish these tasks as
part of their work.
To complete this portrait of our participants, Appendix 4 presents detailed profiles for
each employee. These profiles present their answers to the questions outlined above, and
provide a complete list of their tasks.
54
Task Nb of participants
Participates in operations pertaining to the management of student files. 2
Participates in operations pertaining to the management of human resources. 6
Participates in operations pertaining to the management of financial resources. 8
Participates in operations pertaining to the management of material resources. 8
Follows up on the decisions taken by his/her supervisor and make sure things get done. 13
Manages his/her superior's agenda (fixes appointments according to priorities). 9
Manages the unit's accounts, registers expenses, produces reports, etc. 7
Updates the personnel files and prepares the appropriate documents for hirings, nominations, promotions,
vacations, absences, etc.
5
Accomplishes the student files technician's tasks when needed. 2
Receives and filters phone calls, gives out information. Welcomes visitors and prepares necessary
documents.
8
Sorts incoming mail according to priorities. Prepares responses and takes note of any necessary follow-up. 9
Organizes meetings by sending out invitations, booking meeting rooms and preparing the required
documents. Takes part in meetings, takes notes, and ensures any necessary follow-up.
6
Manages the local filing system and applies the required archival procedures. 10
Takes notes and types various documents such as letters, reports, memos, etc. Prepares page layout and
verifies grammar and spelling. Signs certain documents.
13
Participates in the planning of special events such as conferences or travels. 6
Accomplishes the tasks of a similar or lower-ranked position when needed. 8
Table 3.3: List of tasks accomplished by participants.
55
3.4 Validity and trustworthiness in grounded theory
Ensuring the validity of a study means identifying and addressing the potential
methodological weaknesses that may undermine the quality and eventual generalizability
of the results obtained (Maxwell, 1996). In this section, we first look at how validity is
evaluated within both the quantitative and qualitative paradigms. We then present other
criteria developed especially for use with grounded theory.
In the quantitative realm, the traditional criteria are internal and external validity,
reliability and objectivity (Trochim, 2006). Internal validity means that the causal
relationship between an independent and a dependant variable is adequately demonstrated.
External validity is the degree to which the results can be generalized and applied to other
members of the population studied. Reliability refers to consistency in measurement, with
regard to the instruments used to gather data. Finally, objectivity means that any source of
subjectivity, or personal point of view, is eliminated. This can be partially achieved by
having multiple people observing the same phenomenon (Lincoln & Guba, 1985).
While these criteria are usually appropriate in a quantitative paradigm, they may not
readily apply in the qualitative realm, in which our study is situated. For example,
internal validity applies to studies in which the objective is to verify a causal relationship,
which is often not the case in qualitative studies, where the focus is on the observation,
description and discovery of phenomena. It is also impossible to calibrate a researcher
the way an instrument of measure can be in order to ensure reliability. In their 1985 book
entitled “Naturalistic Inquiry”, Lincoln & Guba proposed a new set of alternative validity
criteria more appropriate for qualitative studies. Although somewhat based on the
traditional quantitative criteria, they take into account the peculiarities and concerns of
the qualitative approach. These new criteria were called credibility, transferability,
dependability and confirmability. Lincoln & Guba (1985) also discarded the term validity,
referring instead to the concept of trustworthiness.
56
Credibility means that the results of a study are “credible or believable from the
perspective of the participant in the research” (Trochim, 2006). This is largely attained in
our research by the fact that most of the early theoretical constructs could be tested by
presenting them to participants. Because data collection and theory building are largely
overlapping steps in grounded theory, it is possible to validate the researcher’s insights by
asking participants whether or not they agree with a certain assumption or conclusion on
the researcher’s part. The use of the constant comparative method of data analysis in
grounded theory also acts as a built-in tool for enhancing credibility. Through this
method, "the data consistently guide the conceptual development of the labels and
definitions for categories and properties as well as their interrelationships. Thus, this
validity check is done during the process of generating the grounded theory product
rather than after it is produced" (Atwood, Hinds, Benoliel & Artinian, 1986, p. 144).
Triangulation is another important element to alleviate credibility concerns. In this study,
triangulation was achieved in part through the use of observation in addition to the
interviews. Since the researcher had access to the workers' computer files during the
interview, it was possible in some instances to verify if what was being said by the
participants was indeed implemented and operationalized on their own workstation. This
helped confirm, infirm or at least nuance what the participant was saying.
Additionally, seeing the way documents were organized helped us get a better grasp of
the participants' tacit knowledge (Gourlay, 2002), and brought up aspects of their
behaviour that they might not have mentioned on their own. Our goal was thus to avoid
relying solely on participants' verbal reports of their information management behaviour,
but also to witness it first-hand. This means that we potentially covered more ground and
obtained a more in-depth view than if we had only used phone interviews, for example. It
also helped ensure the reliability of the interview data, or at least alerted us to potential
discrepancies between the participants' discourse and the reality observed, allowing us to
make sense of things right away by asking for a clarification on the spot.
57
Transferability implies that the results of the research are generalizable, that is, applicable
to other contexts beyond the initial study setting (Trochim, 2006). The responsibility for
ensuring transferability rests mostly on the person doing the generalizing, but the initial
researcher can help in that regard by providing rich descriptions of the initial study
setting and sample, which is what we have done here. This helps other researchers
determine the applicability of the research in their own context.
Dependability requires that the researcher account for any changes in the research setting
which could potentially explain changes within the results. This is the equivalent of
reliability in the quantitative paradigm. In this study we have done this in large part by
giving a contextual backdrop to the results presented. For instance, profiles were
established for each participant, describing their main tasks within their work
environment. The reader is therefore able to judge the evidence presented (e.g. quotes
from participants) with regard to the specific conditions in which that participant works.
Dependability is also somewhat related to the reliability of the researcher as the main data
collection instrument in the research. As a researcher involved in a qualitative study, it is
necessary to be aware of any possible bias. It is widely acknowledged that completely
eliminating bias is impossible, so one must strive to identify and embrace bias (Hale,
Treharne & Kitas, 2007; Mehra, 2002). In the conduct of interviews, as well as in the
subsequent analysis phase, a conscious effort was made to avoid focusing only on data
that confirmed our preconceptions and expectations, or data that fit our initial conceptual
framework. Although the literature review presented earlier is useful in situating the
context for our study, it should in no way determine or influence its outcome. It is
difficult to completely avoid this bias, of course, but being aware of the danger is already
an important part of the solution.
The role of the literature has always been a significant point of contention among
grounded theory experts (Seldén, 2005). The myth of the researcher as a blank slate is
unfortunately still associated with grounded theory (Urquhart & Fernandez, 2006), but
the consensus in recent years seems to be that it is usually unrealistic to pretend not
58
knowing what we already know, and that beyond the risk of contaminating data
collection and analysis in the current study, knowledge of previous studies can actually
enhance theoretical sensitivity and thus help the researcher recognize patterns in the data
(Weed, 2009). The main concern remains to avoid rigid preconceptions which might taint
observations and thus the resulting theory (Glaser, 1998). We tend to share this view,
which is anyway much more realistic in the context of a doctoral research project, where
a student is expected to bring into the research a hefty baggage of previous theoretical
knowledge. Vigilance is still necessary, of course, but the idea, as Becker (1986, p. 149)
puts it, is simply to “[u]se the literature, [not] let it use you.”
Finally, confirmability is the extent to which a research’s findings could be corroborated
by others. One way this is addressed in our study is the inclusion of numerous quotes
from participants. Such quotes are the material upon which we were able to draw
conclusions about the topic of study, so readers are able to draw their own conclusions on
the strength of the evidence presented.
Although our study appears to fulfill all of the trustworthiness criteria necessary for
qualitative research, some grounded theory advocates would probably argue that such
criteria should not even be applied to grounded theory, because it does not share the same
ontological and epistemological assumptions (Weed, 2009). In their initial presentation of
grounded theory, Glaser and Strauss (1967) proposed four criteria by which grounded
theory should ultimately be judged: fit, work, relevance and modifiability.
Fit “relates to how closely the concepts and theory generated fit the incidents and
phenomena they represent” (Weed, 2009, p. 506). We can then say that a theory fits the
empirical situation studied. Fit is ensured by the use of the constant comparative method
and by theoretical saturation, which were both employed in our study. Work means that
the developed theory has explanatory power, and can give insights into the reasons for
the phenomena observed in the field. Relevance refers to the fact that the theory deals
with the real concerns of the people involved in the research setting, that it is relevant to
them, and that it addresses some of the core problems and issues they are facing (Giske &
59
Artinian, 2007). Finally, a grounded theory should be modifiable, which means that it
should be open to further iterations, developments and extensions, as additional
discoveries are made through further research. In fact, the grounded theory generated
from a single study cannot claim to be generally applicable; it is merely a substantive
theory, applicable in a more limited, substantive area (Glaser & Strauss, 1967). However,
it can then be combined with the products of other studies so that a more formal theory
may emerge.
60
Chapter 4: Survey of document spaces This chapter presents an account of the observations made during the interviews with our
participants. When participants gave us a tour of their documents, we would notice where
documents had been placed, and would ask questions accordingly. The goal was to better
understand the makeup of our participants’ documentary landscape.
The chapter is organized by document spaces, which correspond to the main locations
where a worker is likely to store documents. As we shall see, although most participants
used similar document spaces, and while there were obviously similar elements in each
case, there was also significant variation with regard to how they used each space. We
have tried to give an idea of the various possibilities we encountered.
In the following pages each type of location will thus be described, with an emphasis on
highlighting the variation observed between participants. In chapter 5, we will focus on
identifying the possible reasons that might explain such variation.
4.1 Defining document spaces
It is tempting to view the documents controlled by workers as a single entity, stored in a
specific location (say, the My Documents folder) on their computer workstation. In reality,
documents tend to be scattered in more than one location, even within a single
workstation. Increasingly, workers also tend to store files in shared folders or intranets.
Every location may serve a different purpose, and be subjected to specific rules regarding
organization, file naming conventions, weeding, etc.
The locations we have identified in our study are: the Windows20 file structure (main and
secondary folders), the Windows Desktop, the e-mail application, paper documents and
finally shared environments. These locations, which we have dubbed document spaces,
20 All participants interviewed used the Windows operating system.
61
will be described in more detail in the present chapter. In each case, we will outline the
main trends observed as well as the exceptions to the rule, and discuss the significant
issues workers have to contend with.
In the following pages, we will also regularly mention the number of participants that
displayed a specific behaviour, expressed a particular feeling or use a particular method.
These numbers, however, should not necessarily be taken at face value. In grounded
theory, it is common for some participants to be asked questions that other participants
are not asked. The reason is that although an interview guide is normally used at the
beginning of the study, new topics of inquiry and discussion are likely to come up as the
study progresses. As such, the interviewer may ask questions about certain topics to later
participants, but not to earlier participants, as this particular aspect had not yet surfaced.
In other cases, a topic is discussed with earlier participants, but not later on, if the
researcher considers that the issue has been sufficiently documented, and that no new
elements are likely to surface. This may seem problematic, but we have to remember that
the goal of grounded theory is not to paint a statistically representative portrait of a
situation, but rather to explore a topic without pre-formed hypotheses and discover new
elements along the way. Therefore, the exact numbers are given here as an indication, but
should not be taken as a precise reflection of how widespread a particular element was
within our sample.
4.2 Windows file structure
We first focus on the files located on the worker’s personal workstation. This was until
recently workers’ main location for storing files. As we shall see later, the advent of
centralized, departmental servers has somewhat reduced the importance of the individual
workstation in some cases. Nevertheless, the personal computer remains a significant part
of every office worker’s information landscape.
When simply asked to show how they had organized their files, most of the participants
in this study instinctively started with the file structure contained in Windows, with the
62
exception of those who were storing their files on a shared server. And although workers
can theoretically put files anywhere on their workstation, some definite trends were found
in that regard.
4.2.1 Main folder
All workers appeared to have one main location for their work-related files: a base folder
which contains a more or less developed hierarchical file structure. This main folder was
often Windows’ My Documents folder, or a subfolder within My Documents. This is not
very surprising, as that is generally where new Windows users are advised to keep their
files. More importantly, it is often the default saving location for files in Windows,
notably when working with applications of the Microsoft Office suite (Word, Excel,
PowerPoint, etc.). These applications are widespread in the workplace, and were used
exclusively in our studied environment.
The main folder’s name often corresponded to the worker’s name (e.g. “Francine”). In
other cases the main folder was created at the root of the hard drive, e.g. C:\. The main
folder was normally divided into a number of subfolders, each likely to contain several
subfolders and files, and so on.
The main folder usually contains most, if not all of the files used in the course of daily
work, whether produced by the worker or saved after having been downloaded or
received via e-mail. These are the files that are usually given to the successor after
someone leaves a position, or to their temporary replacement in the case of an interim.
These are also the files that are most likely to be shared with (or at least accessible to)
colleagues.
Overwhelmingly, the workers in this study were completely free to organize their files as
they pleased within this main folder. They had the freedom to create new sub-folders as
needed, for example when a new project or responsibility surfaced. Unsurprisingly, there
63
were significant differences among participants in the way this folder was organized.
This is somewhat expected, because despite having a similar job description (say, office
work technician), participants were employed in various departments and services, and
thus had relatively dissimilar responsibilities and topics of concern.
There were a few exceptions to this, however:
Three workers were using the University’s institutional classification scheme,
called. This scheme consists of a set of rules which prescribes first- and second-
level subdivisions. It was conceived so that it could theoretically be used by any
unit within the University. Obviously, workers using this common system would
have similar folders and sub-folders within their main folder. At lower levels of
the structure however, the classification scheme does allow some freedom in
organization, which means that some additional folders can be created, as long as
one respects certain rules (i.e. new folders may only be created within certain
prescribed folders, but not in others).
Some workers shared their documents with their supervisor. In that case, the main
folder was common to both people, and was usually located on a centralized,
departmental server.
In the above two situations, there was of course much less freedom for workers to
organize things as they pleased, as they had to deal with additional constraints. The
specific effects of such constraints will be described in greater detail in the next chapter.
4.2.1.1 Characteristics of the file folder structure
In the PIM literature, file folder structures have often been characterized by the number
of hierarchical levels they comprise (depth) and the number of folders present at each
level (breadth). These variables have been the object of several studies in the past, such
64
as Gonçalves & Jorge (2003), Khoo et al. (2007) and Henderson (2009) just to take a few
examples. Statistics such as the average number of folders per level can be interesting,
and certainly the above-mentioned studies have provided useful insights, but in the end
such statistics cannot tell the whole story, and thus can sometimes be misleading, unless
there is a certain level of uniformity in the participants’ situations.
For instance, how can we reasonably compare the structure of an employee who has held
the same position for over 10 years and one who has just started in a position? Yet this
would have been the case in our study, since our theoretical sampling meant comparing
employees in different context (experienced veterans, new employees, etc.). Therefore, it
would have been rather difficult to draw general conclusions on our participants’
behaviours or preferences by looking at aggregated data such as the average number of
first-level folders, the total number of folders or documents, etc.
Adding to the difficulty in our study is the fact that in several cases, the folder structure
we observed was not solely a creation of the participant we met. They had often inherited
their structure from their predecessor, so the observable structure could in fact be the
amalgamated result of several successive employees adding documents through the years,
sometimes modifying parts of the structure along the way. Therefore, in most cases, such
quantitative data wouldn’t explain much about the rationale behind a person’s file
structure, hence our decision to focus on interviews to understand these aspects and to
focus on observing finer points of the structure rather than drawing conclusions based on
a superficial view of the structure as a whole.
That being said, while we didn’t compile extensive data about the characteristics of the
structures we encountered, we did make a few observations in that regard. But the
emphasis was clearly put on asking users to comment on the structure they had developed,
in order to try and understand why it looked the way it did. This sometimes included
questions about specific aspects of their structure such as depth or breadth, which resulted
in the following observations.
65
First of all, some of our participants said they preferred a shallow structure, where
everything would be available in as few clicks as possible. One participant felt that
having to click more than two or three times to get to a document was too time-
consuming, and thus had consciously developed a much shallower, flatter structure.
Others preferred having more hierarchical levels, because then they didn’t have to search
and/or think as much on each level to find what they were looking for. They felt more
“guided” as they browsed through their files. For example, one participant had taken over
a position where documents had been classified into 25 folders at the first level of the
hierarchy. She had then undertaken a complete overhaul of the structure, and in the end
there were only three folders left at the first level. To her, this made a lot more sense,
especially since her responsibilities could be roughly separated into three categories
anyway. When looking for a file, she was able to navigate to the appropriate category
from the outset, thus eliminating navigation in the other two thirds of her documents. An
added motivation for the change was that she didn’t like having to scroll up and down to
see everything, as 25 folders wouldn’t fit the entire screen.
In the end, depth and breadth appeared to be largely a matter of personal preference and
efficiency. Workers primarily organize their structure in a way they feel makes their life
easier. Unfortunately, participants didn’t seem to have great insights as to why they
preferred one way or another. They simply did. The question seemed to come down to
clicks vs. cognitive load (Jones et al., 2005). A deeper structure reduces cognitive load, as
there are less options to click on at each level of the hierarchy. On the other hand, more
clicks can get tiresome and be perceived as a waste of time.
However, we must also consider that even though participants state a preference when
asked to choose between a shallow or a deep structure, that personal preference alone
does not ensure that the structure will look a certain way. In some cases, the complexity
of a position simply calls for a more detailed hierarchy, even though the worker could
have done without that many levels based on personal taste alone. So personal preference,
66
although it undoubtedly plays a big part, is probably not the only determinant of file
structure morphology.
4.2.1.2 Main organizational criteria
In terms of organizational logic, most workers’ file structure followed a predominantly
thematic organization. Documents were organized by topics ranging from general areas
of responsibility (“Accounting”, “Inventory”) 21, specific projects or events (“Fundraising
Dinner 2008”), departments or specific people such as clients or colleagues (“Dupuis,
Catherine” or “Assistant Dean”), etc. All of these would normally be mixed up at the first
and level of the hierarchy and in subsequent levels as well. Type of document (“Annual
reports” or “Expense accounts”) was also quite common.
On the second level, folders could again be divided according to themes, but things often
followed a temporal logic, with folders for years and/or months. This was prevalent for
pretty much every participant we interviewed, which may have to do with the rather
cyclical nature of the academic environment. Indeed, workers tend to perform similar
tasks year in and year out (preparing the recruitment campaign, welcoming the new
students, preparing the class schedule, publishing the annual report, etc.). Therefore, it
makes sense to group together all the documents related to a specific year by creating a
new folder for them. We were told that this helps avoid confusion between very similar
documents (and very similar filenames), and makes it easier to identify and retrieve all
the documents pertaining to a particular year.
A third mode of organization was observed in this study, not among our participants, but
in the folders left behind by some of our participants’ predecessors. On two occasions,
the previous employees had organized all of their files by software. There would typically
be a “Word” folder, an “Excel” folder, etc. Inside these folders, files would follow a
21 The folder names given here are fictitious, but are representative of actual folders observed during our
study.
67
normal thematic organization. It was assumed by our interviewees that these workers,
usually people in their fifties or sixties when they started using a PC, had first learned to
use Word years ago, and had then put every Word file they created in a Word folder; the
same with Excel, and so on.
We were unfortunately unable to contact these workers, who were now retired, to try and
understand the rationale behind their organizational scheme. What we can say, however,
is that such a structure can create significant problems for their successors, who have to
try and guess the format when looking for a document. Luckily, in one of those cases, the
new employee rarely needed to refer to her predecessor’s files.
4.2.1.3 Orphan Files
Documents are normally located in folders. Although they do serve a purpose in
establishing the logic of the structure by acting as thematic guideposts, the main function
of folders is still as containers for files. Nevertheless, not every file fits easily inside a
sub-folder, which means that within a folder, some independent, residual files may
remain. We call these orphan files. A fictitious example is presented in Figure 4.122.
22 Microsoft product screen shots reprinted with permission from Microsoft Corporation.
See http://www.microsoft.com/canada/permission/default.mspx for details.
68
Figure 4.1: “bcrp.xlsx” and “Meeting 2008-05-26.docx” are orphan files, as they
remain outside of the existing sub-folders.
Orphan files may not seem like a significant issue, but they were a source of concern for
at least five of our participants, who viewed them as evidence of a messy or faulty file
structure. When encountered along the way during a tour of their file structure,
participants would often apologize for their presence, saying that these files “normally
shouldn’t be there”23 or that they just hadn’t had “time to clean up”24. They sometimes
appeared to feel genuinely guilty about their presence, and seemed afraid we would judge
their document organization performance in large part on the presence of such files. A
similar phenomenon was previously mentioned by Jones et al. (2005) as well as
Boardman & Sasse (2004). In the latter study, roughly a third of the participants actually
tidied up their documents during the interview.
23 “Ces fichiers-là, normalement ça ne devrait pas être là.” (Jeanne)
24 “Euh… ça tu vois c’est juste que c’est moi qui n’a pas eu le temps de faire le ménage” (Anne)
69
Orphan files may exist for a few different reasons. First, they may not a have a clear
destination within one of the existing sub-folders. Alternatively, someone may have just
put the file there because they were in a hurry, and didn’t want to bother about deciding
in which folder to put them. Whatever the case may be, they remain in limbo until a
cleanup occurs and they are either deleted, assigned to an existing folder, or a new folder
is created to welcome them:
I like to have sub-folders. When there’s more than three documents related to
the same topic, I like to create a little folder and put them inside. Three, that’s
pretty much my limit. I don’t like when things are just lying around. Maybe
I’m crazy [laughs].25
In some cases however, their presence is entirely voluntary:
These are things I like to have on hand, that I didn’t put in folders. They
could be in a folder as well but… I find that having them there makes it
easier… […] I’m used to having them there. 26
In that case, quick and easy access is considered more important than a so-called “tidy”
file structure. Nevertheless, most participants seemed to view orphan files as a problem to
be addressed, which brings us to the issue of “cleaning up”. This activity appears to be
universal among all of our participants, although it can take various forms. It is mostly
done on an ad hoc basis, when stumbling upon a file that shouldn’t be where it is. Action
is then immediately taken to address the situation. It can also involve simply going 25 “Moi j’aime ça avoir des sous-répertoires. Quand il y a plus que trois documents qui ont rapport au
même sujet, j’aime ça créer une petite chemise pour les mettre dedans. Ma limite c’est à peu près ça, trois.
Je n’aime pas ça quand les choses traînent. Je suis peut-être folle ! [rires]” (Lise)
26 “Ça c’est des choses que j’aime avoir à portée de la main. Ils pourraient être dans un répertoire aussi
mais… je trouve que de les avoir là ça facilite… […] Je suis habituée de les avoir là.” (Béatrice)
70
through the structure and correcting problems as they are spotted. Some try to do this
once in a while when time permits, but such operations tend to take a backseat to regular
tasks. Others will do the bulk of it about once a year, as a sort of spring cleaning activity
(actually done in the summer, when there is usually less to do). And some prefer to apply
what could be called preemptive cleanup, which is the habit of always saving files in a
proper folder, creating it on the fly if necessary.
4.2.1.4 Dealing with multiple versions
The habit of keeping several versions of a document was widespread among participants,
and although it can become tedious to sift through numerous versions of the same
document, it does bring some measure of reassurance to know that it’s possible to go
back in time and recuperate a missing piece of information, or simply get back to the
previous state of a document. Even those who are somewhat overwhelmed by the number
of versions tend to see the bright side:
Versions… Boy do we have versions! You see, Final version, Version 1,
Version 2, Version 3… I'm not necessarily in favor of this. Well, there's the
good and the bad. On the one hand it allows you to go back and see what you
did before. On the other hand there is the possibility of mistakes. 27
The main potential problem here involves using or sending a version one assumes is the
latest when it is, in fact, not. And the probability of error increases significantly when two
or more people collaborate on a single document, sending it to each other via e-mail.
27 “Les versions ! On en a-tu des versions ? On en a ! Tu vois, version finale, version 1, version 2, version
3. Moi je ne suis pas nécessairement favorable à ça. Ben il y a le pour et le contre. Oui si tu veux reculer en
arrière puis voir ce que tu as fait antérieurement. Mais non parce que il y a possibilité d’erreur.” (Anne)
71
Although some appear to keep previous versions indefinitely, others make a point of
deleting all but the final version once a project is completed.
Researcher: But once the project is completed, you do a big cleanup.28
Participant: Probably that in the end, in November, when everything is settled,
then I’m going to delete. I’ll remove what’s in the way, and just keep the final
[version]. 29
4.2.1.5 Naming files and folders
Unless they use the university’s institutional scheme, which as we have said was not very
often the case, our interviewees had total freedom to determine the name of their files and
folders. This ends up being both a blessing and a curse. The freedom is certainly
appreciated, but such freedom can translate into problems for their successors.
The main issue is quite simply that the names chosen are not always very clear or
representative of the contents. The use of abbreviations can contribute to this, but there is
also the fact that naming a file or folder, especially when done in a hurry, may sometimes
be an afterthought.
A second problem is the lack of uniformity from one document to the other. Some of the
participants did not seem to care too much about uniformity or consistency in file names.
For instance, when one participant was asked why about half her folders had titles in
capital letters and the other half in lower case letters, she said that she didn’t have a good 28 “Mais une fois que le projet est fini, tu fais un grand nettoyage.” (Anne)
29 “Probablement qu’à la fin, au mois de novembre, quand tout va être réglé, là je vais effacer. Je vais
enlever ce qui va me nuire, je vais juste garder le final.” (Béatrice)
72
reason and actually didn’t mind at all. But for others this is a very important issue, even if
the information is still somewhat visible in an inconsistent naming scheme. Here is one of
those workers, describing the files her predecessor left for her:
P: You see, it's never written the same way. Myself, I try to present the
information in a uniform manner. When it's “Classes Winter 2008”, “Classes
Winter 2009”, etc., it's not going to be Winter completely written out one
time, and then just W the next, like this. It's not clear!
R: So you try to standardize file names?
P: Systematically. I need that in order to find my way around.30
Apart from standardization, the general rule of thumb is to try and convey as much
information about the document as possible, so that “you don’t need to open the file to
know what’s inside.”31 Of course, a more specific filename should also help distinguish
one document from the next.
Common practices among our participants include adding various metadata directly to
filenames. Dates are regularly added (e.g. the fiscal year, or the exact date of a meeting),
but other keywords are used as well, for example the names of guests present at a specific
meeting. Such additions can be helpful when browsing a folder in search of a specific
document. It can also be useful when conducting a search query by file title. Once again,
efficiency seemed to be a key driver in developing such initiatives.
30 “P : Tu vois, ce n’est jamais écrit de la même façon. Moi j’essaie d’uniformiser mes informations. Quand
je sais que c’est Cours - Hiver 2008, Cours - Hiver 2009, etc., ça ne va pas être Hiver au complet 2009, des
fois ça va être un H, des fois, comme ça. Ce n’est pas clair !
R : Alors tu essaies de normaliser les noms de fichiers ?
P : Moi c’est systématique. Pour me retrouver moi, j’ai besoin de ça.” (Ginette)
31 “[…] pas besoin d’ouvrir le fichier pour savoir ce qu’il contient” (Béatrice)
73
Sometimes the additions were made in part to help determine the order in which the files
will appear in a folder, thus countering the default file ordering provided by the Windows
software. This was done by adding special characters (e.g. ‘zzzz’, ‘#’ or ‘_’) or even
dates, at the beginning of a file’s name, in order to push it up or down in a list. This was a
way for participants to circumvent some of the sorting limitations inherent to the
Windows interface.
The attention paid to naming files and folders certainly has its uses for the individual
worker, but some also have a decidedly altruistic attitude about the whole thing:
I’d like to build a system where anyone looking for a file could just sit here
and find it, that would be my ideal. And when I save a file, that’s what I think
about, I have this preoccupation of… for someone from the outside, how I
could name it to facilitate its retrieval. 32
Obviously not all participants displayed such an attitude, but it was surprisingly
prevalent. We will deal with this point in the next chapter.
4.2.1.6 Searching and browsing
When looking for a file, users can adopt one of two strategies: searching or browsing
(Barreau & Nardi, 1995). Searching involves the use of a search engine (the built-in
Microsoft Windows search engine in our case), while browsing means navigating through
the various files and folders until arriving at the desired location/document.
32 “J’aimerais ça monter un système qui fait en sorte que n’importe qui cherche un dossier s’assoie ici puis
il puisse le retracer, ce serait ça mon idéal. Et quand je sauvegarde un fichier, c’est à ça que je pense, j’ai
cette préoccupation de… pour quelqu’un de l’extérieur, comment je pourrais l’intituler pour faciliter son
repérage.” (Béatrice)
74
Participants in our study were not especially enthusiastic about using the search engine,
but most did use it from time to time. Search is basically never employed for files they
use on a regular basis, and rarely for files that they have created themselves, especially if
they were created recently. In such cases, browsing to the file is by far the preferred
access method, as they usually know exactly where it resides.
Searches are mostly done when they suspect that a file exists, but they’re not sure where
it is. In fact, one of the main reasons for using the search engine is when they are looking
for a document located among their predecessor’s files. They are usually not very
familiar with how things are organized there, so browsing would be much more arduous.
This is also true of documents located within their own files, but that they are not overly
familiar with, such as documents they might have received via e-mail but never actually
read. In some situations, their superior might ask them for a specific piece of information
that they are supposed to have, but they have little or no idea where it might be. In that
case, they may choose to search inside the files instead of just in the file titles. That being
said, such searches are not always successful.
When asked why they don’t use the search engine all that much, some simply say that
they’re just “not used to it” or that they plainly “don’t like it”. But most report that
searching is just not the first thing that comes to mind, and that it is normally used as a
last resort, after having browsed for a while.
It should be noted that none of the participants were using one of the more advanced
Desktop Search applications that have been launched in recent years, such as Google
Desktop or Microsoft’s Windows Desktop Search. Perhaps this partly explains their
rather low usage and overall level of satisfaction.
75
4.2.2 Secondary folders
If the main folder is where the bulk of the work files are located, there can also be other
folders located completely outside the main folder hierarchy, which serve a different
purpose. We’ll call these secondary folders. From our observations, there are three main
kinds of secondary folders: personal folders, buffer folders and templates folders.
Personal folders, as their name implies, are generally used to store personal, non work-
related files. This could be pictures of children or pets, humoristic pictures or texts
received from colleagues and friends, class material if the worker is taking classes, or any
other personal documents such as personal letters, an Excel file to keep track of personal
weekly expenses, etc. Workers spend a significant part of their life at work, so it’s not
surprising that their personal life tends to creep up in their computer in the form of
personal files. But a conscious effort is usually made to separate these from the rest of the
files. Personal folders were often identified as such, with names such as “Perso” or
“Personnel” 33.
The second type of secondary folder is buffer folders. These are used to store ongoing
work, before it is transferred to the main folder. Therefore, they contain works-in-
progress or documents at a draft stage. Typical names are “En cours” 34 or “Brouillon” 35.
Why are such folders necessary? For some, the main folder tends to be perceived as a
place where definitive, final versions of documents are stored, and they may feel
uncomfortable putting working versions of a document there. This is especially true in
the case of shared folders. In that case, putting a document in folder equates to publishing
it for all (or at least a limited number of colleagues) to see. It is thus considered
preferable to place a document there only after it has reached a certain degree of quality
or completeness. Before that, it remains in the buffer folder, until it is ready to be
published.
33 French for “Personal”. 34 French for “Ongoing”. 35 French for “Draft”.
76
A third type of secondary folder is the templates folder. As its name implies, this folder
contains examples or models of the most frequent documents an employee may have to
produce. This can go from simple letters to full-fledged MS Access or Excel-based
mailing list systems. Another common example is a contact list, or even a more
developed list of procedures, in the form of a guide or how-to:
P: It’s something I create for myself every time I arrive in a new position […]
it has 11 pages in all. All of it is information about TGDE [Student records
management technician].
R: What is it, rules to follow, codes?
P: Information on how to access administrative applications, how to do
withdrawals, registrations… All the procedures, the people to contact when
you need information. The codes used in applications. How to do
equivalencies, exemptions… And I’ve done this again since I’m in this
department. It’s the type of information I could need anytime.36
The goal here is obviously to save time by reusing tools and tricks developed earlier,
often in previous positions. Of course, this type of document is especially prevalent in the
case of more experienced, well-traveled workers. The templates folder tends to travel
36 “P: C’est quelque chose que je me fais à chaque fois que j’arrive dans un poste […] j’ai 11 pages en tout.
D’informations sur les TGDE.
R: C’est quoi, c’est des règles à suivre, des codes ?
P: Des informations pour aller dans les applications administratives, comment est-ce qu’on fait des
abandons, comment est-ce qu’on fait des inscriptions… Toutes les procédures, les personnes-contact,
quand on a besoin d’info. Les codes utilisés dans les applications. Comment on fait des équivalences, des
exemptions… Et puis j’ai aussi fait la même chose depuis que je suis au département ici. C’est le genre
d’information dont je peux toujours avoir besoin. ” (Ginette)
77
with them when they leave a position. It contains enough general information to be useful
in other similar positions throughout the organization.
In practice, the distinction between personal, buffer and templates folders is not always
clear, which means that a single folder may very well serve all three functions. For
instance, sometimes there is indeed a link with their work, but the documents therein are
not perceived as being directly related, hence the separation:
I’ll use [this folder] only if I’m currently doing something or for more
personal things, honestly. For example, I’m a member of the Social
Committee, so I put that in there. I did something for Friends of [the
University], but it wasn’t necessarily for my normal work… I was also a
student [at this University] for a while, so I put that there as well.37
Another participant mentioned her involvement in one of the University’s yearly charity
events, and stored the related documents in a secondary folder as well. The idea seems to
be that even though these tasks are done during work hours or are somewhat related to
their workplace persona, they represent tasks or responsibilities undertaken on an
individual, voluntary basis, but not really as part of their normal job description. As such,
these documents shouldn’t be mixed with the purely work-related documents.
The simultaneous existence and use of such totally separate folders may appear
unnecessarily cumbersome, but this separation, though it may seem a bit artificial, is
usually there for a reason, and the workers interviewed wouldn't necessarily have it any
other way:
37 “Moi je vais utiliser [ce répertoire-là] juste si j’ai quelque chose que je suis en train de faire ou des
choses plutôt personnelles, honnêtement. Comme, tu vois, je suis membre du Comité social, je mets ça là.
J’ai fait quelque chose pour Friends of [the University] mais c’était pas nécessairement pour mon travail
régulier… Là à ce moment-là c’était comme personnel. J’ai aussi été étudiante à [cette université] pendant
un bout de temps, ça fait que je mettais ça là.” (Carole)
78
R: It seems that there are three very distinct spaces in your computer. The
first is the folder you share with your boss, the second is “En cours”
[“Ongoing”], and the third in “My Documents”. It seems that you're always
moving between those three spaces.
P: Most of the time, yes.
R: Have you ever thought of grouping all this in the same folder? For you,
there is a logical demarcation between those three?
P: There is indeed a demarcation. In “My Documents” it's my own documents.
In “Ongoing”, it's related to the office, it's what is ongoing but not necessarily
finished. And the things in the shared folder, well they're more finished, and
it has to do with her [points to her boss's office]. For me there's really... it's
really demarcated like that.
R: OK, so the things in “Ongoing” are things you're currently working on,
and when you're ready to share it with her it's like... you make it public in...
you publish it.
P: You could put it that way, yes. Or if it's kind of off-topic, but could be
useful, like the itinerary to go to a conference. I wouldn't go ahead and put
that in her things [i.e. the shared folder]. I'll put them in “Ongoing”. The
correspondence related to the leaflets, I'm not going to give that to her right
away, because I don't see the use for now. Eventually, when it's finished, then
yes, but not right now.
R: But if, for instance, you were to put the itineraries in the shared folder now,
would there be a disadvantage to that?
79
P: I don't know. I could have put them there. But I would have had to create a
new a category and I don't feel like it - there are already enough categories as
it is... [laughs] 38
Whatever the case may be, secondary folders tend to have a more private character,
which partly explains why they are located apart from the main folder. As such, they are
never shared with other workers and tend to be deleted when an employee leaves a
position, especially in the case of a personal folder.
38 “R: On dirait qu’il y a trois espaces vraiment distincts dans votre ordinateur. Premièrement il y a le
dossier partagé avec votre patronne, le deuxième c’est “En cours”, puis le troisième c’est “Mes documents”.
On dirait que vous allez et venez toujours un peu entre ces trois espaces-là.
P: La plupart du temps, oui.
R: Est-ce que ça vous avez déjà pensé à regrouper tout ça dans le même répertoire ? Pour vous il y a une
démarcation logique entre ces trois-là ?
P: Il y en a une démarcation. Dans Mes documents c’est mes documents à moi. Dans En cours c’est par
rapport au bureau, puis c’est ce qui en cours mais pas nécessairement fini. Puis ce qui est dans Dossier
commun, ben là c’est plus fini, puis c’est avec elle [elle pointe vers le bureau de sa patronne]. Pour moi il y
a vraiment… c’est vraiment démarqué comme ça.
R: OK, alors ce qui est dans En cours vous êtes en train de travailler dessus, puis quand vous êtes prête à
partager avec elle, là c’est comme… vous le rendez public dans… vous le publiez.
P: On pourrait dire ça comme ça, oui. Ou si ça n’a pas rapport, mais ça peut être utile, tu sais comme les
itinéraires pour se rendre à une conférence. Je n’irais pas les mettre dans ses affaires à elle. Je vais les
mettre dans En cours. Les correspondances à propos des dépliants , je ne lui donnerai pas tout de suite à
elle, pcq je ne vois pas l’utilité pour le moment. Éventuellement, quand ça va être terminé, bon, oui, mais
pas pour le moment.
R: Mais si vous mettiez, par exemple les itinéraires tout de suite dans le dossier partagé, est-ce qu’il y aurait
un désavantage à ça ?
P: Je ne sais pas… J’aurais pu mettre ça là. Mais c’est parce qu'il aurait fallu que je crée une autre catégorie.
Puis ça ne me tente pas, il y en a déjà assez... [rires]” (Diane)
80
4.2.3 Windows Desktop
Generally speaking, participants’ Windows Desktop (WD) was not very cluttered. The
most common component found was shortcuts to software applications, which were used
by all participants. This is not very surprising, as even a default installation of the
Windows software will usually include Desktop shortcuts to the Recycle Bin, the My
Documents folder, etc. In this case, the university’s IT department had also created a few
default shortcuts for all their workstations, for instance to Internet Explorer. The workers
interviewed had usually added additional shortcuts to Microsoft Office applications, or to
more specialized applications (financial or accounting modules, student records
applications and so on). Ravasio, Schär & Kruger (2004) observed similar uses.
Document files also tended to be stored on the WD, usually on a much more temporary
basis, and for a variety of specific reasons, which we outline below.
4.2.3.1 Templates and reference documents
We mentioned the use of a templates folder earlier, but in some cases, templates or other
similar documents are simply left on the WD, especially if they are used on a regular
basis:
P: This is the standard PowerPoint template for the University. I leave it there
so it’s easily accessible.
R: It’s something you use frequently?
81
P: Oh yes. My boss is always making presentations left and right. Like, every
week or so. So I just reuse the same template as a starting point. It’s like a
reference document for me.39
If the above document is likely to stay on the WD on a permanent basis, most other
documents appear to stay there for a much shorter period.
4.2.3.2 Documents currently being worked on
If some like to use a buffer folder for ongoing work, others simply leave such documents
on the WD, especially if the work is likely to be completed within a few days.
If my boss tells me “Make a table to summarize this or that”, you can be sure
I’m not going to put it in [our shared folder] right away. I’ll put it on my
Desktop. Only on the Desktop, that’s the way I like to work. I put it on the
Desktop until it’s been approved, and when it’s really official, then I’m going
to transfer it to [the shared folder].40
39 “P: Ça c’est le modèle PowerPoint standard de l’Université. Je le laisse là pour qu’il soit
facilement accessible.
R: C’est quelque chose que vous utilisez fréquemment ?
P: Oh oui. Ma patronne est toujours en train de donner des présentations PowerPoint à gauche puis à droite.
Genre, à chaque semaine. Ça fait que je réutilise toujours le même gabarit comme point de départ. C’est
comme une sorte de document de référence pour moi. ” (Anne)
40 “P : Si ma patronne me dit « Fais-moi un tableau qui récapitule telle ou telle chose », c’est sûr et certain
que je ne le mettrai pas tout de suite sur [le répertoire partagé]. Je vais le mettre sur mon Bureau. Juste sur
mon Bureau, moi c’est comme ça que je fonctionne. Je mets ça sur mon bureau juste le temps que ce soit
approuvé, puis quand c’est vraiment plus officiel, bien là je vais aller le transférer sur [le répertoire
partagé]. ” (Ingrid)
82
In the following example, files are simply destroyed after their stay on the WD, since
they are of no further use after that:
P: Here you’re going to see on my Desktop, the little documents floating
around are things I only keep for a day or two… I also take care of the
website, so if someone sends me a picture to put on the website, I save it on
my Desktop because I have to work on it, crop it, change the format, etc. So I
keep it there. And when I’m done, when it’s on the site, I just delete, I clean
up, I remove it.
R: Could there be a case where you would just move it somewhere else?
P: Um, no… No, normally if I put it there, it’s because it will be there only
for a very short time… I don’t even think “I’m going to put it in that folder
so I can find it later”. I know these are files I won’t have to keep.41
In other cases we were told that files were likely to stay there for up to a month, but not
much longer. The files are then classified where appropriate in the hierarchy of the main
folder, where they’ll be available if needed.
41 “P: Ici vous allez voir sur mon bureau, les petits documents qui traînent c’est que c’est des trucs que je
conserve une journée ou deux… Je m’occupe aussi du site Internet, alors si on m’envoie une photo à mettre
sur le site Internet, alors je la sauvegarde sur le bureau parce que j’ai à la retravailler, la recadrer, changer le
format, et tout ça. Alors je la garde là. Et puis quand c’est fait, c’est sur le site, bien je supprime, je fais le
ménage, j’en enlève.
R: Est-ce que ça peut être juste que vous allez classer ça ailleurs ?
P: Euh non… Non, normalement si je le mets là, c’est que ça va être là vraiment de courte durée… je ne
réfléchis même pas à bon… « je vais le classer dans tel répertoire pour faciliter la recherche ». Je sais que
c’est des dossiers que je n’aurai pas à garder.” (Béatrice)
83
4.2.3.3 Documents in transit
Another reason to put files on the WD is that it can be very quickly and easily accessed
when saving or sending a file. In Windows software, the WD is often suggested as one of
several possible saving locations, if not the default location. As such, it often becomes
people’s default saving location, even though files will likely be moved to another
location soon afterwards.
The same applies when choosing the file location of an attachment when writing an e-
mail (see figure 4.2). Because of this, one of our participants has developed the habit of
copying any file she wanted to send as e-mail attachment on the WD first, and then
sending that copy of the file. For her, this is cognitively much easier than having to guide
Outlook through the whole Windows structure in order to specify the file’s location.
Once the e-mail has been sent, the Desktop copy is simply deleted.
Another participant had a rather nonconventional use of the WD, which deserves further
attention. Anytime she needed to work on a file located somewhere in the hierarchy of
her main folder, she would create a shortcut to it on the WD and then open this shortcut.
After having done the required modifications, she would simply delete the shortcut to the
file. That way the document would remain classified while it was being worked on.
She explained that this was akin to what she had always done with paper documents: she
would take a document out of the file cabinet to work on it, and then put it back in there
after she was done. This method made sense to her in a paper environment, so she applied
a similar way of functioning when she started working in a digital environment. In her
mind, Word was simply the digital equivalent of a typewriter, and all else followed suit.
This certainly goes to show how a person’s mental model can influence the way software
is used, and it highlights the transitory, temporary nature of the WD as a file storage
location.
84
Figure 4.2: The Desktop is one of the few suggested locations when adding an
attachment to an e-mail.
4.2.3.4 Reminding
The reminding function of the Windows Desktop has been reported by several
researchers in the past, such as Malone (1983) with the physical desktop table, or Barreau
& Nardi (1995) in its virtual form. This was confirmed in our study as well, as five
participants mentioned it. An example, among others:
P: These are travel expenses. As long as they’re not reimbursed, I leave them
there [on the WD]. But once they’re paid, I delete them, you know.
R: The documents stay there so you’ll remember to follow up?
85
P: Yes, that’s exactly why.42
4.2.3.5 Messiness
Most of the participants interviewed were very reluctant to have too many elements,
documents or otherwise, on the WD. Some viewed it as equivalent to having a messy
(physical) office or desk, something a professional secretary or similar worker should
avoid at all costs (“I hate having a messy Desktop. It’s just like my physical office!”43).
Some displayed a certain pride in how neat their Desktop was, while others said they felt
guilty because there were “more things than usual” there, or they hadn’t had the time to
“clean up” like they “usually do” before my visit.
4.3 E-mail
Another important component of a worker’s information environment is e-mail. Without
exception, participants in this study used the Microsoft Outlook software. It includes,
among other things, Contacts and Calendar modules.
As in the case of the Windows file structure, workers usually enjoy full freedom in
organizing their e-mail. Any number of folders and sub-folders can be created, and e-
mails can be distributed at will within these folders. In four cases the secretary had access
and managed her superior’s e-mail, in addition to hers. The reason for this was explained
by one participant:
42 “P: Ça c’est des frais de déplacement, de voyage. Tant qu’il n’est pas payé, je les laisse là. Mais un coup
qu’il est payé, je le mets dans la corbeille, tu sais.
R: Les documents restent là pour vous rappeler de faire le suivi ?
P: Oui, c’est en plein pour ça.” (Édith) 43 “J’ai horreur d’avoir un bureau chargé. C’est comme mon bureau physique !” (Ginette)
86
It’s because we were receiving a lot of duplicate e-mails, or, especially for
external organizations, [my boss] would receive it. Sometimes I was on their
list, and sometimes I wasn’t. So he didn’t know if I had received the e-mail or
not, if he had to forward it to me. Because nowadays people often use hidden
lists! [i.e.: putting a list of e-mails in blind carbon copy] […] In some cases it
was “Thanks for sending it to me, I hadn’t received it.”, and other times I
would just delete it. “Did you get it?” “Yes, yes, I did.” So after a while it
became confusing.44
4.3.1 Weeding
With the quantity of e-mail that circulates in the modern workplace, e-mail applications
can get crowded pretty quickly, and are held at least partly responsible for the
phenomenon of information overload (Whittaker & Sidner, 1996). That being said, very
few participants in our study had specific weeding rules or guidelines. By default, most e-
mails were kept just to be on the safe side, unless they were seen as truly irrelevant to a
person’s responsibilities. It was however quite common to keep only the last e-mail in a
conversation, as this e-mail would usually include a copy of previous exchanges. All
previous messages in the conversation could thus be safely deleted.
One participant did have a more radical approach, but she remains an exception within
the group:
44 “C’est parce qu’on recevait beaucoup de courriels en double, ou sinon, surtout pour les organismes
externes, [mon patron] en recevait. Parfois j’étais sur leur liste, et parfois je ne l’étais pas. Alors là il ne
savait plus si le courriel je l’avais eu ou je ne l’avais pas eu, s’il fallait qu’il me le fasse suivre. Parce que là
les gens sont partis souvent à se faire des listes cachées ! […] Puis il y en a c’était “ OK, merci de me
l’avoir envoyé, je ne l’ai pas eu ”, puis il y en a d’autres ben je les effaçais. “ L’as-tu eu ?” “Bien oui, bien
oui, je l’ai eu.” Alors là un moment donné ça devenait mélangeant.” (Édith)
87
P: [Speaking about her e-mail…] Once a year I delete everything though. I
clean up and delete everything. I clean out almost the entire year, except the
last few months.
R: And that isn’t kept anywhere?
P: No. That’s it. In the end, these are not very special documents. If it were
really correspondence, then correspondence is printed and placed in the file
cabinet. If there’s a follow-up, it’s printed. Or stapled with the document in
question if it’s related to an order. This is so that we can follow up, so that
there will be something if someone comes back five years later. 45
All of the participants we interviewed would print at least some of their received e-mails,
but the vast majority would usually also keep the digital version, “just in case”. In this
worker’s mind, one copy was sufficient, and the possibility to do a good cleanup of her e-
mail was just too appealing. The other workers interviewed seemed to go by the
philosophy that “two copies are better than one.” We will deal with the phenomenon of
printing e-mails in section 4.4, dedicated to paper documents.
4.3.2 Attachments
There doesn’t seem to be any steadfast rules when saving attachments in the file folder
structure, although every participant seemed to do it to some extent. In fact, sometimes
45 “P : [À propos de ses courriels…] Une fois par année je vide tout par exemple. Je fais un ménage et je
vide tout. Je vide l’année quasiment entière, sauf les derniers mois.
R : Et ce n’est pas gardé nulle part ?
P : Non. That’s it. Dans le fond ce ne sont pas de grands documents. Si c’était vraiment une correspondance,
la correspondance est imprimée et placée dans le dossier. S’il y a une suite à avoir, c’est imprimé. Ou c’est
broché avec le document en question, si c’est une commande. Pour qu’il y ait un suivi, si qqn revient 5 ans
en arrière, qu’il y ait qqch.” (Manon)
88
the attachments are printed, but then left in the e-mail. In the end, attachments are really
evaluated on a case by case basis:
R: So you don’t systematically save the attachments?
P: No. Not systematically. Depending on the degree of importance, or if
we’re going to use it… if it’s something we’re going to work on.46
According to our participants, an attachment is more likely to be saved if it seems like an
important document – a document one wants to make sure they keep, because they might
have to refer to it later – or because they have they work on, modify or enhance that
document in some way. In the latter case, it is obviously necessary to save the document
in question in order to be able to make any changes to it.
Some go a step further by saving entire individual e-mails in the Windows structure. In
such cases, selected e-mails related to a project will be included in that project’s folder,
along with other types of files (PDF, Word, etc.). The goal is simply to have a single
place where all the information about a topic can be found. This idea of keeping
everything together is an important trend observed in this study, one we will also come
back to in the next section on paper documents.
An additional advantage of keeping e-mails in the file structure is that while file folders
are often shared between an employee and his/her superior, e-mail is much less likely to
be - at least from what we observed in our sample. If both keep separate, password-
protected mailboxes, saving key e-mails in a shared file structure ensures that both can
access the information when necessary. Two participants mentioned employing this very
method in order to ensure the availability of crucial e-mails.
46 “R: Donc tu ne vas pas sauvegarder systématiquement les pièces jointes ?
P: Non. Pas systématiquement. Selon le degré d’importance ou si on va l’utiliser… si c’est quelque chose
sur lequel on va travailler.” (Anne)
89
4.3.3 Correspondence between Windows folders and e-mail
There was usually not a perfect correspondence between e-mail and Windows folders (or
paper documents, for that matter) in terms of their logical organization. Although this
was often seen as an ideal by workers, it was usually considered infeasible in practice.
Three participants mentioned that they had tried for a while to keep both areas in perfect
sync but they had never quite succeeded, and they ended up abandoning the idea. They
were satisfied with having a rough correspondence between the two, without necessarily
striving for perfection.
The initial goal was to have the Outlook classification be the same as in the
physical file cabinets. In the beginning anyway, and we do try to keep up, but
I must say that there are small differences, it’s not always 100%. It’s not
always easy. It’s just that sometimes it’s harder. It’s feasible, but it really
takes… how should I put this? Like you see this one right here is not exactly
the same as this one here. That’s because I wanted to highlight this one,
because it was a special campaign, so I left it outside. Anyway, we’re not that
far, I’d say 85%. 47
This cross-tool disparity is not entirely surprising, given the complexity and the amount
of e-mails and files that workers have to manage. This issue was even studied in depth by
Boardman et al. (2003). That being said, perfectly synchronized cross-tool organization
might not be adequate for everybody, and one of the four participants in Boardman et
al.’s (2003) study preferred having the flexibility to organize things differently in each 47 “P : Le but initial c’était d’avoir la classification du Outlook la même que dans les classeurs physiques.
Au début en tout cas, mais on essaie de garder ça mais je dois dire qu’il y a des petites nuances, c’est pas
toujours 100%. C’est pas toujours facile. C’est que des fois c’est difficile. C’est faisable, mais ça prend
vraiment… comment je pourrais dire ça ? Comme tu vois ça ici, on n’a pas ça nécessairement la même
chose. C’est parce que je voulais faire ressortir celui-là, parce que c’est une campagne spéciale, alors je l’ai
laissé à l’extérieur. En tout cas on est pas loin, je pourrais dire à 85%.” (Anne)
90
environment. In that study, using special software, Windows, e-mail and Web bookmark
folders were automatically mirrored or users were prompted to mirror them in each
environment.
One important consideration is that e-mail is first and foremost a communication medium,
which seems to influence folder organization. It was quite common in our study to see e-
mail folders named after contacts, something that was much less frequent in the case of
Windows folders, the names of which tended to refer primarily to responsibilities or
projects. So even though there is unavoidable and significant overlap between
environments, the one size fits all approach to classification might not always be the best.
Although cross-tool uniformity is far from being the norm, there was one significant
exception in our study. One manager had taken upon himself to craft a classification
scheme for his service, which for the moment was only used by him and his secretary,
who I was able to interview. In this case, a two- or three-letter code identified every
folder in Windows, Outlook and the file cabinets. One admitted goal of this endeavor was
to ensure that even if his assistant left (and this particular position had been affected by a
lot of personnel turnover through the years), the document organization would remain
consistent. We will discuss this aspect in more detail in the next chapter.
4.4 Paper documents
Despite numerous predictions about the advent of the so-called paperless office, it is hard
to deny that paper is still very present and used regularly in many a workplace, and
certainly in those we visited over the course of our study. In fact, it would appear that
paper and electronic documents are likely to coexist for quite a while.
Every worker interviewed maintained at least one filing cabinet, but some had much
more than that. In one particularly extreme case, a participant’s predecessor had been in
her position for over 20 years and had the habit of keeping “everything”. Her paper
91
document collection occupied several filing cabinets and a few supply cabinets located in
two additional separate rooms. In her office, almost every surface was covered by a huge
pile of paper. When that employee retired and her replacement (our participant) started
cleaning up the piles of paper, other employees would compliment her on her new pieces
of furniture. They were then shocked to learn that the furniture had always been there, but
had literally been buried under mounds of paper for all these years.
Thankfully, not every case is that extreme, but paper does remain present in - if not
central to - the conduct of daily activities. Although the workers interviewed tended to
acknowledge the advantages of electronic documents, notably the possibility of full-text
searching through the content, there are some things for which there is simply no digital
equivalent yet, or at least not to a level as satisfying or convenient as paper.
First, we must remember that many documents are not born-digital, but rather exist
throughout their lifetime in paper form. Purchase orders, bills, contracts, checks and
official letters are examples of very common documents which are received and kept in
paper form, and such documents are rarely if ever digitized to allow their inclusion
among a worker’s electronic documents. These documents may have to be kept in their
original format for several years for accounting purposes, or because they feature the
original signature.
A recurring theme in interviews is the possibility of constituting a complete file for a
project or client in paper form, something that is hard to do in digital form. Of course,
one can put all the (digital) files related to a topic in the same (digital) folder, but paper
documents are still considered much easier to use.
It’s because if she needs to study a file, she has everything at hand. Like the
other day, she had brought a file home. She needed to remind herself of what
that project was all about. It’s easier to take the pile of papers, the file, and
92
shuffle through it as you please, than to go through a lot of Word documents
and then click, click… Now she has the physical thing.48
Even within a single folder, information in a computer is often scattered into several
distinct files, which usually have to be opened one by one in order to see their contents.
On the other hand, a paper folder can be a seamless mix of documents of various origins
(contributed by various employees) or formats (documents originally in paper form, as
well as printed e-mails, PDF files, text from Word, tables from Excel, etc.).
The paper folder thus tends to become the go-to place to get all the information on a
particular project or topic. In turn, this encourages employees to continue printing
documents and adding to the file in question. Given that some documents will always be
in paper form anyway, it seems logical to print the rest in order to have a complete file:
My reflex is always to print the e-mail and include it in the folder. Unless it’s
really useless, but otherwise, if it can add something to the file, I’ll print it
and add it. So that the file is complete. That way, if someone ever needs that
file, they can see all the e-mail exchanges right away. 49
An additional benefit of having such a complete file in paper form is that their superior
can easily pick it up and bring it to a meeting, without missing any key information. Four
workers reported printing most of their documents for this exact reason.
48 “C’est parce que si elle a besoin de consulter un dossier, bien là elle a tout en main. Comme l’autre jour,
elle avait apporté un dossier chez elle. Elle avait besoin de se remémorer ce qui s’était passé dans ce
dossier-là. C’est plus facile de prendre la pile de papiers, le dossier, puis de feuilleter ça comme tu veux
que d’aller voir plein de fichier Word un par un puis cliquer, cliquer… Là elle l’a physiquement.” (Diane)
49 “Mon réflexe c’est toujours d’imprimer le courriel puis de l’inclure au dossier. À moins que ce soit
vraiment inutile, mais sinon, si ça peut apporter quelque chose au dossier, je vais l’imprimer puis je vais
l’ajouter. Pour que le dossier soit complet. Comme ça si jamais quelqu’un a besoin du dossier, elle peut
voir tout de suite tous les échanges de courriel.” (Ingrid)
93
For some reason, paper also tends to be perceived as more final, more definitive. Three
workers told us that they mostly printed the final versions of documents pertaining to a
project, once that project was completed. In a way, it’s as if printing a document put the
finishing touch, and finalizes the matter. Perhaps this is due to a perception that paper
documents are somehow more legally binding or are expected to last longer than digital
documents. Four participants also mentioned that they would print documents they
consider especially useful or important, documents which contain information they want
to have on hand at all times, for quick reference.
Within a department or work unit, printed e-mails and documents can eventually form a
sort of paper knowledge base that is accessible to all employees, much more so than if
electronic documents were located in individual workers’ file structures or e-mail inboxes.
For instance, e-mail accounts are usually password-protected, and are likely to follow an
employee when they leave a position. The material that ends up being printed tends to
represent a good sample of the documents produced by an organization. This knowledge
base can thus represent a very useful resource for new employees, by providing templates
for the most common documents they may have to create. One very recently hired
employee mentioned that looking at existing examples had been helpful when the time
came for her to produce similar documents.
Paper can also have a significant reminding function, as was first highlighted by Malone
(1983). One participant showed us several e-mails she had printed and intentionally left
on her desk. Each e-mail was annotated with a few words summarizing the required
follow-up. To her, these printed e-mails served as excellent reminders of things to do.
She could have simply left these e-mails in her inbox, but she found that follow ups were
more difficult that way. E-mails tend to get marked as read even when she does not want
them to, so in the end it is much clearer for her to simply print them. The printed e-mails
are usually not kept after a task is done, so they have a very temporary purpose. “Perhaps
these are the old, well-established habits of an old secretary”, she said50. 50 “C’est peut-être parce que j’ai des habitudes bien ancrées de vieille secrétaire… ” (Kim)
94
Interestingly, three participants mentioned environmental concerns when talking about
paper documents. It would appear that recent efforts to raise awareness of over-printing
(such as “Do you really need to print this e-mail?” e-mail signatures) are starting to come
to fruition. They admitted feeling a bit guilty about printing in recent years, sometimes
now thinking twice before they did. Still, even if such considerations are on their mind,
not everyone is looking forward to a paperless environment:
I know there’s a tendency to print less and less, and we often get reminded of
this. “If it’s not necessary to print this…” But I find it very hard. I told [my
boss], if one day there is a strict rule not to print anymore, it will be at your
own risk! 51
In the end, paper still has many advantages which are likely to delay the advent of an all-
digital workplace. New technologies might change things in the future, but in the current
landscape, it does appear rather unlikely. It should be said that some of the workers
interviewed did want to see paper disappear, or at least become less prominent in the
workplace, but they also understood that it might not necessarily be possible in the short
term, as the habit of printing is probably too ingrained in people’s minds and in current
processes and workflows.
4.5 Shared environments
One important trend we noticed in our study is the increasing prevalence of shared
environments (e.g. shared folders). This is opposed to the classic model where every
51 “Je sais qu’il y a une tendance de moins en moins imprimer, puis on se le fait rappeler souvent. Si ce
n’est pas important d’imprimer… Mais j’ai beaucoup de difficulté. J’ai dit à [ma patronne], si un jour il y a
une règle stricte de ne plus imprimer, c’est à vos risques et périls.” (Kim)
95
person manages his/her files on his/her own workstation, in somewhat of a closed circuit.
Obviously it is always possible to transfer files via e-mail, a CD-ROM or a USB key, but
the idea here is that dedicated, common environments are created, where people can
access each other’s files.
Sharing documents this way can be especially interesting for an office worker and his/her
superior, who tend to work hand in hand on the same documents. This section will thus
focus more specifically on this type of working relationship, which is especially relevant
in our study of support staff, and was quite common among our participants.
In our study, the following scenarios were observed:
Completely separate file structures: Every worker has his/her own individual
computer workstation, with no sharing mechanisms. No link between the
employee’s and the superior’s documents.
Access rights: Access is given, to one or several colleagues, to a folder located on
someone’s workstation. A worker can share a folder from his/her own structure
by right-clicking on the folder’s properties and modifying access rights and
security settings, thus allowing colleagues to access that folder. The superior can
grant his/her employee access to his/her files, and vice versa.
Centralized data located on a departmental server: In this model, every worker’s
main folder resides on the same server. Typically, every employee will have
his/her own individual directory on the server, and may or may not have access to
their colleagues’, depending on the access rights granted. A typical example of
this involves an employee and his/her superior working in the same shared folder,
which resides on the server. They can both still maintain secondary folders on
their respective computers, but their primary work files will reside on the server.
One major advantage of the server solution is that a single backup procedure can
be used to save an entire department’s files, while the two decentralized models
96
mentioned above require every worker to perform a backup procedure on his/her
own computer.
As for e-mail, here are the sharing possibilities we observed:
Completely separate e-mail: The employee and the superior each have their own,
distinct e-mail accounts.
Superior’s e-mail managed by employee: Here, both the employee and the
superior may have their own e-mail account, but the employee can access the
superior’s e-mail and perform any required task there. In some cases, the
employee always has access, but will only actively manage the account (replying,
deleting, etc.) when the superior is on holiday or absent for an extended period.
Collective e-mail account: This type of e-mail account is meant to have several
owners. The idea is that instead of having personal accounts, people will share
another, common account, which can be given a more neutral name. For example,
a department’s director and his/her assistant may share an account called
administration@... E-mails sent to this address can be managed by both people.
For the employee/superior duo, shared environments present significant advantages. The
main improvement might be that it puts a stop to the incessant sending of documents via
e-mail between the two. Documents are left in the shared folder and modified there,
instead of being exchanged every time a new version is produced. E-mail attachments are
thus drastically reduced, and above all, workers report much less confusion among
existing versions. Furthermore, documents are available at all times, even when someone
is away from the office.
However, shared environments do present some drawbacks, especially from the
assistant’s perspective. The main reason is that the shared folder will usually be
organized from the perspective of the superior, who decides what the structure will
97
ultimately look like. For the office worker, not being able to make such decisions on
his/her own can become a little frustrating. Working in a shared folder can also be a
source of confusion. For example, someone can decide to classify a document into a
given folder, but the other may expect to find it elsewhere. Alternatively, both can decide
to put a file in different folders, while ignoring that the other has already done so.
That being said, our interviewees didn’t seem too troubled by these issues, which they
seemed to view as relatively minor disagreements, certainly something they could live
with:
R: Do you sometimes find it limiting to have to follow an already established
structure? DO you sometimes tell yourself: "I wouldn't have put that there, I
would have done things differently"?
P: Sometimes there are little moments of frustration, I must say. But after that
I tell myself it's not a big deal. [...] The thing is it also depends on your
mental structure. If I think it's more useful to put it somewhere else, but the
person wants to have it there... In that case I'll adapt, you know? But there's
still... It can be a little frustrating.52
52 “R : Est-ce que vous trouvez ça limitant d’avoir à vous conformer à une structure déjà établie? Est-ce
que des fois vous vous dites : “J’aurais pas vu ça là, j’aurais fait ça autrement…” ?
P : Des fois il y a des petits moments de frustration je dois dire. Mais après ça je me dis, c’est pas grave. [...]
C’est parce que ça dépend de notre structure mentale aussi. Si moi je le vois plus pertinent ailleurs, mais la
personne veut l’avoir là... À ce moment-là je vais m’adapter, tu comprends? Mais il y a quand même… il
peut y avoir une frustration.” (Anne)
98
4.5.1 Coordination mechanisms
Secretaries and their superiors use various coordination mechanisms to ensure that things
go as smoothly as possible when working in a shared environment. Sometimes the
superior will give very specific instructions as to where to put a file (e.g. “Can you file
this under X?”).
When secretaries are uncertain of where they should put a file or whether a new folder
should be created, they can of course simply ask. However, because the superior is not
always available, and because they might not want to bother him/her too often, some tend
to keep aside the few files they’re unsure of, and eventually sit down with their boss to
decide where they should go.
Some workers also reported that they often made suggestions to their superior, whether
about where to store a particular file, or about the file structure in general (creation of a
folder, merger of existing folders, etc.). The superior will usually have the last word in
such situations, but it does give support staff a chance to have their say, and makes this a
more collaborative effort, rather than a unilateral decision process:
Since I work in close collaboration with [my boss], I often consult her. When
there are several classification possibilities. And sometimes we look into it
and then we decide that for this specific type of situation, we think it would
be better to put it in this folder, in that place. And we come to an agreement
because we've thought about it together.53
All in all, good communication was seen by participants as a key factor in the success of
shared environments.
53 "Comme je travaille étroitement avec [ma patronne], souvent je vais la consulter. Quand il y a plusieurs
possibilités de classement. Puis des fois on approfondit le sujet puis là on se dit, pour telle situation bien
précise, on pense qu’on serait mieux de le classer dans tel dossier, à tel endroit. Puis là on se met d’accord
parce qu'on a réfléchit ensemble." (Kim)
99
100
Four of the workers we met expected that Microsoft SharePoint would soon be
implemented in their department in the coming months or years. SharePoint is a
document management platform that includes collaborative functions. It was not in use at
the time of our study, but it is worth mentioning here as it signals a definite trend, one
which might have interesting implications for document management in the workplace.
101
6B116BChapter 5: Factors and causes of variations
In the previous chapter, we described the document spaces used by our interviewees, and
some of the related behaviours they had developed over the years. These employees share
relatively similar job descriptions, and work in the same large organization. However,
although there were some aspects that were common to all participants, a more prevalent
tendency was to have significant differences in the ways workers handled their
documents. It therefore becomes interesting to wonder what factors might lead to such
variation.
It is of course very tempting to try and generalize by stating that office support staff
workers as a group always do things in such or such a way. Unfortunately, the reality is
considerably more complex. This leads us to try and determine why such variation occurs.
Every worker’s behaviour is likely dictated or rather influenced by a significant number
of factors. To a certain extent, every worker's situation is somewhat different, not only in
terms of the context in which they work, but also in terms of their own individual
preferences, their level of experience, the influence of their superior, etc.
Over the course of our interviews, we were able to identify several factors that can
contribute to shaping a worker's document space in a specific way. In this chapter, we
identify seven different groups of factors and show the impact they can have on a
worker’s PIM behaviours.
The factors gathered here were identified after a thorough procedure of coding, memo-
writing and conceptualization, using the constant comparative method. We were
eventually able to isolate these seven categories of factors from the mass of data gathered.
In many cases, we have also chosen to illustrate the various factors and their properties
by using quotations from participants. These quotations are presented here as additional
evidence, but also for their explanatory power; we felt that in such cases the participants’
102
own words would convey meaning more effectively and succinctly than the researcher’s
explanations ever could.
29B139B5.1 Job Content
The first set of factors can be grouped under the umbrella of job content. In studies such
as this one, it is tempting to focus on a specific type of position (e.g. engineers, professors
or office clerks), and then paint a portrait of the typical worker. However, when we look
more closely, every specific position within the organization has its own peculiarities. In
other words, two office work coordination technicians at this university may have the
same job description, but certain elements can make their actual work very different. This,
in turn, is likely to influence document organization. In a sense then, every position is
unique, and such discrepancies should not be ignored.
83B193B5.1.1 Breadth and complexity
The first element to consider is the relative complexity or breadth of a worker’s tasks and
responsibilities. Some positions are considered more complex than others, because of the
number of projects in which the employees are likely to be involved, the number of
contacts they may interact with on a daily or yearly basis, etc.
For example, high-level managers, such as the director of a department or service, are
usually required to have a global view of all the activities taking place in their sector.
Although their involvement in every ongoing project is likely to be minimal, they have to
at least be aware of them, as they may be required to make decisions about them. As a
result, the director’s personal assistant will have to handle documents about all these
various projects, which could potentially mean a broader document organization, with
more first-level categories in the classification scheme. On the other hand, a person with
a narrower scope of tasks, who is mostly involved in one or two major projects, or has
just a few core tasks, may use less first-level folders. This is an example where job
103
content directly affects the shape of the file structure. For example, in our study, Édith,
Jeanne and Kim acted as personal assistants of directors, and they all had markedly
developed, extensive file structures, when compared to some of the other participants we
interviewed, such as Administration Technicians (AT). One AT had been an Office Work
Technician (OWT) in the past, and could see a difference in the extent of the breadth of
the structure she now had to deal with:
In the other positions it wasn't necessarily like this. Here it's very structured.
But it's because it's not... the scope of the things I have to do is not that wide.
It doesn't go in every direction. It's really focused on administration, budgets...
For administration technicians here it's really... it's pretty much the priority.F
54
The number of contacts a worker has can also be a factor. Some workers may have to
interact with a fairly limited number of people on a typical day, say a few immediate
colleagues. But others may have to deal with dozens, even hundreds of students,
providers, donors, etc. in the course of a year. The need to keep at least some information
on each one is likely to result in a large number of documents and directories, although
not necessarily at the first level of the hierarchy, since the information tends to be fairly
uniform. For instance all of this information could be contained in a single “Clients”
directoy. This is something we observed in our study with Carole, Diane and Ingrid. In
such cases, documents about contacts are kept because they help supplement one’s
memory, for example by providing biographical information about individual donors, or
by keeping track of exchanges and transactions with specific vendors. This is simply
essential in order for these employees to do their job properly.
Similarly, positions with a strong accounting component tend to force one to keep
significantly more documents, often in their original paper form, as these may be needed,
54 “Dans les autres postes ce n’était pas nécessairement comme ça. Ici c’est quand même très structuré.
Mais c’est parce que ce n’est pas… moi en tout cas ce que j’ai à faire ce n’est pas si large que ça. Je ne
m’en vais pas dans toutes les directions. C’est vraiment concentré sur l’administration, les budgets… Les
techniciennes en administration ici c’est vraiment… c’est pas mal la priorité.” (Ingrid)
104
even after several years, for auditing purposes. This was the case for Ingrid, Jeanne and
Manon in our study.
What all of this means is that the amount of documents workers keep within their files, or
the number of folders they create, are probably not entirely attributable to personal
preference, but are likely also due in some part to the specific requirements of the
position. Therefore, the number of folders in a person’s computer only tells one part of
the story. It is crucial to understand why so many folders are kept. Otherwise, that person
is likely to be considered an eccentric or just a statistical anomaly, when they in fact
represent a real and probably not-so-unique case.
84B194B5.1.2 Work Pace
Work pace is another important aspect. Some workers rarely if ever have to deal with
emergencies, while others have to be ready at all times, for example to provide a piece of
information to their superior when they ask for it:
For me, it’s easier to put it in there because it makes it easier to find
afterwards. With me, it has to go fast. I do everything I can to make sure it
does, because I don’t want to have to look for things. […] Because if [my
boss] asks me something, he really wants to know right away. He doesn’t
have time to wait. That way, I can find things easily. F
55
Someone in that situation will usually pay more attention to how easy things are to access.
Some documents may be placed in a more easily available position in the file structure
55 “C’est plus facile de les mettre là-dedans pour moi parce que c’est plus facile à retracer par après. Moi,
il faut que ça aille vite. Je prends les moyens pour que ça aille vite, pour justement que je n’aie pas à tout
chercher. […] Parce que si [mon patron] me demande quelque chose, il veut vraiment le savoir tout de suite.
Il n’a pas le temps d’attendre. Comme ça, je le retrouve tout de suite.” (Jeanne)
105
because they are accessed frequently, even though pure thematic logic would dictate that
they be buried somewhere deeper in the hierarchy. Some workers used shortcuts to link
to such documents. In short, when one expects to have to deliver information on very
short notice, they are likely to take this into consideration and organize their documents
accordingly.
That being said, a quicker work pace can also affect document organization negatively.
When things have to be done quickly, a worker may pay less attention to things such as
saving location, classification, file naming, etc. This may of course create problems later
on. As we saw in the section on orphan files earlier, such files are often left there because
there was no time to decide where else to put them. If a worker is in a hurry, they may
leave a file in a temporary location and plan to return later and give it a proper home,
which may or may not ever happen. Being in a hurry may also bring someone to
precipitate decisions, make classificatory mistakes, drag a document in the wrong folder,
etc.
We should also consider that generally speaking, if someone is very busy, their main task
is likely to overshadow considerations of document organization, which usually means
less time to clean up and address issues such as orphan files:
I often think about it, I always tell myself “I should be doing that, I should be
reorganizing”, but because I don’t have the time, and it’s not a priority, you
know… […] There are times when I’d feel like doing that, but I just don’t
have the time. Then again, when I do have the time it doesn’t mean I feel like
it… F
56
56 “J’y pense souvent, je me dis toujours “Il faudrait que je le fasse, il faudrait que je réorganise”, mais
comme je n’ai pas le temps, puis ce n’est pas une priorité, tu sais... [...] Des fois... quand ça me tenterait de
le faire, je n’ai pas le temps. Puis quand j’ai le temps, ça ne veut pas dire que ça me tente...” (Diane)
106
This confirms Jones’ (2007, 46) conclusion that “people do not generally take time out of
a busy day to assess their organizations or their PIM practice in general.”
85B195B5.1.3 Cyclical nature of the work
Another aspect of a worker’s job is how consistent or predictable it is. As we have
mentioned before, universities are rather cyclical in their functioning, which encourages
workers to organize per school/fiscal year, but also to expect certain tasks to be done or
certain types of information to be requested at certain times during the year. This is
obviously truer for some positions than others, but when it is the case, this can help
anticipate future needs, and thus help guide document organization for better efficiency.
For example, two workers said they knew they would work a lot around budget
preparation time, but a bit less during the rest of the year. For others, two or three major
events during the year proved especially challenging or time-consuming.
One participant told us that although the first year in a specific position can be
challenging, once at least one full cycle has passed (i.e. the Fall, Winter and
Spring/Summer terms, and then back to Fall), one has a pretty good idea of what the job
entails, and can then prepare accordingly. This cyclical nature also means that employees
can refer to a previous year’s folder to obtain a template or at least a good example of
how things were done last year, and thus know what to do this year.
Normally, in your files, you always have your past, which is the year before,
and the current year. So if you’re wondering about this type of bill, you know
you’re going to find one from last year, and then you can see how it was
processed. […] Once you’ve had the one-year cycle, you’ve faced pretty
much every possible situation. After that, it’s much easier. F
57
57 “Normalement, dans tes dossiers, tu as toujours ton passé, l’année d’avant, puis l’année courante. Donc
si tu te poses des questions sur ce genre de facture-là, tu sais que tu vas en trouver une de l’année passée et
107
Because of this, a person’s document structure may be rather tentative during the first
year, but will eventually be adjusted to the realities of the position, which in the end tend
to be relatively similar year after year. This doesn’t preclude the possibility of special
projects or events along the way, but the bulk of the work should at least become more
predictable, which allows for the establishment of a stable, efficient file structure, one
that is well-tailored to the job.
86B196B5.1.4 Domain and context
Another potential obstacle, especially for newcomers, is the use of specialized vocabulary
within a department or service. It can take a bit of time to familiarize oneself with the
terms employed in a given workplace: topics, people, committees, meetings, associations,
forms, projects, services offered and so on.
Support staff usually occupy relatively similar positions whatever department they might
be in. The general nature of their tasks thus remains relatively consistent, but from one
environment to the next, the topics discussed and the nature of the documents exchanged
can vary considerably. For example, working for the director of the Department of
Physics is very different from working in the University’s IT department, or in its legal
department. In all cases, an adaptation and learning period will undoubtedly be necessary.
Three workers reported not always being comfortable having to work with, and
especially classify, documents of which they don’t necessarily understand the content or
meaning. It also makes it much harder to gauge the relative importance of a document. Of
course, such knowledge and ability will come with experience, but it is certainly not easy
tu peux voir comment elle a été traitée. […] Une fois que tu as eu le cycle d’un an, tu as fais face à pas mal
toutes les situations possibles. Après ça, ça va mieux.” (Manon)
108
at first, and in some cases it remains hard even after a year or two, as in the case of this
participant, who worked in the University Libraries:
Yes, I understand the domain better, after two years, but there are still a lot of
things that to me are… I don’t have training in library science, so there are a
lot of things that… Sometimes they talk to each other in terms and My God,
it’s a whole other language… a library language! […] In the beginning I had
some trouble because I didn’t really know the domain, and I find it hard,
when you don’t know a domain, to classify things properly. And it’s not
something you do just like that, getting to know a domain. So over time, I
created folders, without really knowing. There are probably things that could
have been put in another folder, or in a sub-folder, so now I have some
cleaning up to do. F
58
This employee was a very experienced one, but had worked most of her career outside
academia. By her own admission, this probably made it even harder for her, as
universities tend to be rather complex organizations, with a special fondness for
acronyms of all sorts.
58 “Oui, je connais mieux le domaine, depuis mes 2 ans, mais il y a encore beaucoup de choses qui me
sont… je n’ai pas la formation en bibliothéconomie, il y a beaucoup de choses que… Des fois ils se parlent
en termes et mon dieu, c’est un langage… de bibliothèques! […] Au début j’étais un petit peu mal prise
parce que je ne connaissais pas beaucoup le domaine, puis je trouve ça difficile, quand on ne connait pas un
domaine, de bien classer ses affaires. Puis ça ne se fait pas comme ça, bien connaître le domaine. Alors
avec le temps, j’ai ouvert des dossiers, sans trop savoir. Il y a peut-être des choses qui auraient pu être
mises dans un autre dossier, ou dans un sous-dossier, alors j’ai un gros ménage à faire.” (Kim)
109
30B140B5.2 Job status
Aside from the nature of the job description, another crucial aspect with regard to
document management is job status, that is, whether the worker is there on a permanent
or temporary basis.
This seemed to be an especially important factor at the studied university, where the
presence of a Workers’ Union creates very high mobility. When a position becomes
available, the university has to offer it to its own employees before seeking candidates
from the outside. Any current employee can apply for these positions, provided that they
have the required minimum qualifications. Workers have nothing to lose by applying for
a position; their name is simply put on a list ordered purely by years of service. The
person with the most years of service almost automatically gets the job, unless they
decline it, in which case the next person on the list gets called, and so on. It is not rare to
see dozens of people apply for the most coveted positions. So if a person is qualified as
an office work coordination technician, they can potentially work in dozens of
departments and services within the university. In fact, three of the workers we
interviewed had gone through about ten positions in a 20-plus year career, and they did
not consider their situation to be out of the ordinary. Others had had just as many in a
much shorter career, but those were mostly briefer replacement stints.
When asked why they had moved around so much, workers spoke of needing a new
challenge, needing to “keep the passion alive”. They might have been bored in a job they
simply found uninteresting. Some just liked to change jobs once in a while to avoid
boredom. Despite requiring similar qualifications on paper, one job can end up being
very simulating while another can consist mostly of making photocopies… People also
left jobs because they wanted better working conditions (nicer office space, more flexible
working hours), because they didn’t get along with their current boss, or to be reunited
with friends working there. Whatever the reason, making such a move presents very little
risk for a worker, since they get to keep the same salary and years of service, and they
can even come back to their previous job if the new one is not satisfying, provided they
110
do so within 45 days. This explains why position changes are such a common occurrence
at this university and other similar organizations.
Another aspect of mobility is replacements. A temporary worker can be called on to
replace someone who is on maternity leave or sick leave. Some positions may also be
given to temporary personnel on an interim basis, until someone who is fully qualified
takes it up permanently. Many will spend several years going from replacement to
replacement, from temporary contract to temporary contract, especially at the beginning
of their career, until they amass enough years of service, move up in the lists and finally
get an opportunity to obtain a permanent position.
But what does all this have to do with document organization? A lot, actually, because it
turns out that people may act differently whether they are in a permanent position or just
there for a shorter period:
When you’re in a temporary position, and you know you’re not there for long,
often the things you’ll do… sure, you’re going to do it… but… you don’t
really care that much… you know you’re not staying. Whereas me, in a
permanent position like this, I know I’m going to have to work with these
documents and I’ll be careful because I may need them later, you know... The
person who was here temporarily before me was probably very good, but she
didn’t have the same interest that I have. Because I know I’m going to stay.
So I want it to work, and I don’t want to have to rack my brains. I know
where I’m going. So I think it’s worth putting the effort in the beginning,
because I’m going to benefit from this for a while afterwards. F
59
59 “Quand tu es sur un poste temporaire, que tu sais que tu n’es pas là pour longtemps, souvent les tâches
que tu vas faire… tu vas le faire là… mais… tu t’en fous un peu… tu sais que tu t’en vas. Tandis que moi,
un poste permanent comme ça, je sais que je vais avoir affaire à travailler mes documents puis à faire
attention à mes documents pour plus tard, tu sais… Par exemple, la personne qui était ici temporairement
était sûrement très bonne, mais elle n’a pas le même intérêt que moi j’ai. Parce que moi je sais que je vais
rester. Alors là moi je veux que ça marche, et je ne veux pas me casser la tête. Je sais où je m’en vais. Alors
111
However, not everybody will necessarily exhibit the same attitude in such a situation, and
some aren’t afraid to give it their all, even if they know they’re not staying:
Yes, there are certain people that can be influenced by that. For example one
of my friends went through this. She wasn't staying in that job. In the
beginning she wanted to reorganize everything properly, but she didn't want
to give her all for a project she might not even finish. But I don't mind that
kind of thing. It has happened to me, and sure, it's a shame, but it's part of the
game, you know. [...] And some people are just perfectionists anyway. Even
if they stay for just one or two months they want things to be well organized.
So that the next person can have something interesting when she arrives.
Because they're going to know you were there before. In the end it depends,
it's a question of personality. It's all about... the feeling of a job well done. F
60
Clearly, the temporary nature of a position can potentially discourage two common
endeavors of recently-hired workers. The first one is the enhancement of the existing
document organization, on a semi-altruistic basis, in order to leave the situation in a
better position than was initially the case. The other is the desire to mold the document
structure to the worker's liking, so that everything will make more sense to them. In both
cases, the knowledge that their stay in a position will be short-lived may prevent a worker
je pense que ça vaut la peine de mettre l’effort au début, parce que je vais en bénéficier pendant longtemps
après ça.” (Carole)
60 “Oui, il y a certaines personnes que ça peut influencer. Comme j’ai une de mes amies qui a vécu ça. Elle,
justement, ne restait pas sur le poste. Au début elle voulait tout réorganiser les documents puis arranger ça
comme il faut, mais ça ne lui tentait pas de se donner dans un projet qu'elle ne finirait pas. Mais moi, ça ne
me dérange pas. Ça m'est arrivé, puis c’est sûr que c’est plate, mais ça fait partie de la game, tu sais. [...]
Puis il y en a qui sont perfectionnistes quand même. Même si elles restent juste un mois ou deux elles
veulent que ce soit bien structuré. Pour que la prochaine, justement, qu'elle ait quelque chose d’intéressant
quand elle va arriver. Parce qu'elles vont savoir que c'est toi qui étais là avant. Dans le fond ça dépend, c'est
une question de personnalité. C'est juste... le souci du travail bien fait. ” (Jeanne)
112
from undertaking major changes, or putting the necessary effort for an extensive overhaul.
But once a worker is given the reassurance and stability of a permanent position, they can
proceed with the necessary changes without wondering if the effort will be in vain.
31B141B5.3 Existing documents
Unless one arrives in a newly-created position, and in a previously nonexistent
department, workers usually do not start with a clean slate: there are usually some
existing documents left by their predecessors, or simply accumulated through the years
by the department in question. One of the challenges of arriving in a new position is thus
to deal with these existing documents. The amount of existing documents will vary
depending on the age of the unit (whether it’s an old or a newly-established department),
but also on the predecessors’ propensity to weed inconsequential files.
Beyond quantity, the organization of the documents is also important. The predecessor
probably devised her own system, which made sense to her, and which now has to be
decrypted by the newcomer. Also, given the usually high turnover ratio for support staff
positions, the existing documents are likely to have been accumulated by several different
employees over the years, and it is quite common for many if not every predecessor to
have used their own structure, in a separate folder, instead of continuing what was done
previously. As a result, a new employee may have to go through several distinct folder
structures in order to locate a document created before his/her tenure.
For new employees, navigating through the existing documents is always somewhat of a
challenge, and invariably requires a more or less lengthy period of adaptation. Adaptation
is really the key word here, one that was consistently used by our participants. When
moving from one job to another, one needs to adapt to different contexts and different
document structures. For example, a replacement worker may have to use someone else’s
structure during the interim period: “Of course you have to be able to adapt to change.
113
And adapt also in the sense that maybe I wouldn’t have done things that way, but you
have to adapt to what’s there. ”F
61
The level of the adaptation required is likely to vary according to the situation. Not every
file structure is alike, and some seem to be better- or more-logically constructed than
others, which makes life easier for those who have to use it:
Honestly, I didn’t find it too hard here, precisely because it’s well organized.
There are places where…Because in doing replacements you see all sorts of
things… It’s not always easy. But I must say that it’s pretty well organized
here. […] In doing replacements, what I’ve seen is that every person has
their own system. It probably shouldn’t be that way, but that’s how it is. But
it’s pretty rare that I can’t find my way at all. You eventually understand the
logic behind their classification. F
62
Generally speaking, it does help to have at least some experience working in a similar
environment. For instance, academic departments are always somewhat similar to one
another, so experience obtained by working in the Department of Physics would probably
be very helpful in the Department of Philosophy, even though they deal with very
different disciplines. The processes and types of documents involved should still be
similar.
61 “C’est sûr qu’il faut être capable de s’adapter au changement. Puis s’adapter aussi dans le sens que moi
je n’aurais peut-être pas fait ça comme ça, mais il faut que tu t’adaptes avec ce qui est là. ” (Ingrid)
62 “Franchement, ici je n’ai pas trouvé ça trop difficile parce que justement c’est bien organisé. Il y a des
places où… Parce qu’en faisant du remplacement je vois un petit peu toutes sortes de choses… C’est pas
toujours facile. Mais ici je dois dire que c’est très bien organisé. […] Moi en remplacement ce que j’ai vu
c’est que chaque personne a son système… Ça ne devrait pas là, mais c’est ça pareil. Mais c’est assez rare
que je ne me retrouve pas du tout. Tu en viens à comprendre la logique de leur classement. ” (Anne)
114
P: I think it gets better as you gain experience. Since I know the university a
bit more now... When I got into my first position, it was pretty chaotic I
would say...
R: What was particularly difficult?
P: I had no indications. I really had to find things by myself, to teach myself,
like “OK, that's how it works”. I didn't know where the documents were, I
had no instructions, I was not really guided. Even in the computer it was not...
I felt that the filenames were not very representative. [...] I just had no idea. I
was really lost.F
63
However, even with some experience, participants reported regularly having to “use
deduction” or “make hypotheses” in order to understand their predecessor’s structure or
determine the location of a file. A newly-arrived employee, in her first position at the
university, said that when she had free time, she would just go through the various folders,
one by one, in order to familiarize herself and get a sense of what they contained.
87B197B5.3.1 Help from the predecessor
What compounds the difficulty is that, at least at this university, workers rarely got help
or training from their predecessor. Because of the way positions are filled, when a person
steps into his/her new job, his/her predecessor has likely already left several weeks before,
63 “P: Je pense que ça s'améliore avec l’expérience. Comme je connais plus l’université... Quand je suis
arrivée à mon premier poste, c’était un peu le bordel je te dirais.
R: Qu’est-ce qui était particulièrement difficile ?
P: Je n’avais pas d’indications. Il a fallu vraiment que je trouve les choses par moi-même, que je sois
vraiment autodidacte, dans le sens « OK, ça marche comme ça. » Les documents, je ne savais pas ils
étaient où, j’avais moins d’instructions, j’étais moins dirigée. Dans l’ordinateur même, ce n’était pas… Il
me semble que les noms des fichiers n’étaient pas super représentatifs. [...] Je n’avais aucune idée. J’étais
vraiment perdue.” (Ingrid)
115
if not even longer. It is also not a common practice for employees to come back and help
their replacement, except perhaps if they are still working within the same department or
service. A corollary problem is that the new employee may have to deal with a backlog of
documents to classify because no one was doing the job in the meantime, which certainly
doesn’t ease the acclimatization process.
Even in those rare cases where meeting the predecessor is a possibility, it is not always a
cure-all. One participant reported once having the opportunity to spend four days with her
predecessor, who was retiring. But she found that this was really not enough, and that it
only gave her a very superficial view of things, lacking many details. Of course, most of
these four days was used to discuss the actual tasks that the employee would have to
perform, and only a cursory amount of time was used to talk about the existing
documents - which in this case were especially numerous - and the way they had been
organized.
In some cases, the departing employee may not be able to spend a few days with the
replacement, but leaves an e-mail address or phone number, in case the successor is
unable to locate a file, for example. Some of our participants had indeed contacted their
predecessor, despite some initial reluctance to do so. They had never done it more than
once or twice however, mostly, they claimed, because they didn’t want to bother them.
When they did contact their predecessor, they usually obtained the answer they were
looking for, although one participant told us she had stopped calling because she felt that
her predecessor seemed to derive pleasure and sense of superiority out of helping her.
The office worker didn’t want to give her that satisfaction anymore…
Another issue is the lack of documentation that could give pointers to the replacement.
The structure established by departing workers usually remains undocumented; in a way,
the documents are left to speak for themselves. Some of the participants we met had
created documentation for their own use, but also with the assumption that they might
eventually leave their position. That being said, the bulk of those instructions pertained to
116
other aspects of the job, and not necessarily their document structure: things to remember,
useful contacts, etc.
The impact of employee turnover on knowledge preservation and subsequent productivity
in the organization are well-known (Dalkir, 2010). Since a certain amount of turnover is
probably inevitable, it is necessary to ensure that as little knowledge as possible is lost in
the transition, a process known as continuity management (Beazley et al., 2002). This
typically involves various efforts to preserve the tacit knowledge of employees before
they leave.
An employee’s documents may be considered explicit or recorded knowledge (Field,
2003). However, the way these documents are organized is usually not made explicit at
all. In other words, the logic, the rationale behind this organization can be considered
tacit knowledge, and as such, a certain amount of effort has to be made in order to
preserve it. From our observations in this study, this is rarely done.
Only one of the workers we met had the habit of extensively documenting every aspect of
her work. In fact, she had a long experience within the university, and would regularly
change positions when she needed “a challenge”. Such challenges involved taking a new
position and “straightening it out”, “making it better”. The direr the initial situation, the
more interesting for her, and the more satisfied she was with her accomplishments in the
end. Once satisfied with her work, after having put her system and place and put the
position back on track, she would simply leave to undertake a new challenge as soon as
the opportunity presented itself.
Another interesting, and rather extreme, case was reported by one of our participants. She
had to replace someone who was on sick leave for several months. However, not only did
the predecessor not help her get acquainted with the document structure, she actually
forbade her to consult her documents, even locking the file cabinet for the duration of her
absence. Our participant wasn’t exactly sure why the person had done that, but assumed it
was for fear that she would wreak havoc with her file organization. Luckily – and
117
somewhat surprisingly – the replacement told us that in the end, she never really needed
the documents anyway, as she was mostly given new tasks by her superior, who was also
able to provide a copy of necessary documents.
Most of the workers we interviewed had replaced someone else at some point in their
career, which may explain why most of them said that they did try to think of an eventual
successor when organizing their documents. Some even planned on reviewing their
structure before they would leave, in order to make it as clear and streamlined as possible,
notably by getting rid of the orphan files lying around. One participant mentioned that it
was a constant preoccupation for her, and almost every decision she made was done
thinking not only of an eventual replacement, but also of her immediate colleagues, who
might have to use her documents in her absence. This may be the mark of a rather
altruistic person, but it is also a consequence of someone who has done several
replacements throughout her career, a good illustration of Do unto others as you would
have others do unto you.
88B198B5.3.2 Appropriation strategy
Whatever help they may get during the transition, new workers are eventually faced with
the following decision: keep using (and adding to) the existing document structure, or
create their own structure.
Using the existing structure is probably the most common solution in the case of a shorter,
temporary replacement. That way, when the person comes back, everything will be more
or less as they left it, except for the few files or folders that will have been added during
their absence. That being said, not everyone opts for this solution. Many will prefer
creating a new structure outside of the current one, whether they are starting a permanent
position or are just there for a few months.
118
There are several possible reasons for starting a new structure from scratch. Initially, a
worker may not be comfortable with the existing structure, and may think it’s too
difficult to understand and use properly. The structure may seem illogical or artificial,
and the large number of folders and sub-folders may be overwhelming. Starting with a
clean slate and gradually building a new system ensures that the structure accurately
reflects workers’ thinking process and their experience in the position.
A new structure also helps mark a separation between what happened before and after the
start of a worker’s tenure. While the previous structure might have included many now
useless files and folders related to past projects and responsibilities, the new structure
should be, initially at least, a better reflection of the things the new worker has actually
done or been involved with.
The decision to start over may also depend on the state of the existing structure. If the
current structure makes immediate sense to the new employee, they may decide to use it
as is. For some, however, the decision to start anew is automatic, whatever the existing
structure may be:
P: Me, it’s the first thing I do when I arrive in a new department.
Systematically. I don’t touch anything of what the previous person did. I
create a new, separate folder. Every document I create, that doesn’t come
from the other person, I create in my new directory.
R: In this case, is it because you looked at the existing structure and thought
“I can’t use this!” or…?
P: Oh no, not at all. Never. When I arrive I create everything right away,
without even looking at what the previous person has done. Whatever the
case may be, I always create a new folder like this. […] You know, being a
temp, you start to get used to it. Every time you arrive somewhere, you know
there will be an adaptation period, and it’s not always easy, so if I had to
119
adapt to [someone else’s structure] as well… Please, at least let me do things
the way I want. […] I don’t know how to explain it, but in order for me to be
comfortable, I have to start with my own structure, otherwise I go nuts! F
64
In most cases, creating a new structure or not remains a personal decision of the
employee, but it is not without consequences for the organization as a whole. Given the
high internal mobility at this university, it is not rare for several employees to hold a
particular position in succession over the course of a few years. If each employee starts a
new structure, it can quickly become problematic, as files relating to the same project or
topic may end up being scattered in several places, instead of being consolidated in a
single, coherently organized structure.
Five participants clearly stated that they planned on eventually recuperating their
predecessor’s files, and including them within their new structure, as a way of having
everything in one place. However, even though some had started, most weren’t very far
along. They quickly realized that this was not an easy thing to do, since each file had to
be evaluated and reclassified on an individual basis. One participant describes the
difficulty involved:
64 “P: Moi, c’est la première chose que je fais quand j’arrive dans un nouveau département.
Systématiquement. Je ne touche à rien de ce que la personne précédente a fait. Je me crée un nouveau
dossier à part. Tous les documents que je crée, qui ne viennent pas de l’autre personne, je le crée dans mon
nouveau répertoire.
R: Dans ce cas-ci est-ce que c'est parce que tu a vu ce qu’il y avait comme structure et tu t’es dit “Je ne
peux pas utiliser ça !” ou… ?
P: Ah non, pas du tout. Jamais. Moi j’arrive et je crée tout d’office, sans avoir jamais regardé ce que la
précédente a fait. Peu importe, je me crée toujours un dossier comme ça. [...] Tu sais, en tant que
temporaire, on commence à être habituée. À chaque fois qu’on arrive quelque part, on sait qu’il va y avoir
une certaine adaptation et ce n’est pas toujours évident, alors si en plus on devait s’adapter à [la structure de
quelqu’un d’autre]... S’il vous plaît, laissez-moi au moins ma façon de faire. [...] Je ne sais pas comment
l’expliquer, mais pour que je sois confortable, il faut que je parte avec ma propre organisation, sinon je
capote ! ” (Ginette)
120
The worst thing when you switch jobs is you have to get rid of someone
else’s system. You always have to ask yourself: “Why was she keeping this?”
And that’s when you realize that she was insecure, because documents dating
from 1997… She kept everything! You don’t want to delete too much,
because you tell yourself, is this going to blow up in my face one day, maybe
I’ll need it as an example, etc. So you have to take each document one by one
and look at the date, and then decide what you’ll keep. I can’t really think of
an alternative to doing that. F
65
This is obviously a big undertaking, especially if there is a long history of files to
recuperate. Therefore, in most cases the predecessor’s files will remain as is. When
necessary, the new worker will venture in that folder to locate a file, or may conduct a
keyword search in that directory. In some cases, certain files may be copied into the new
structure and modified as needed, with the original left in place.
32B142B5.4 Relationship with the superior
While most of the participants interviewed answered directly to a single person (typically
someone in middle or high management), the nature of the working relationship was not
always the same. The traditional superior-secretary dyad appears simple enough, but
when we look more closely, we realize that there are many ways in which this
relationship can present itself. This, in turn, will have consequences on the document
65 “Le pire quand tu changes de job c’est qu’il faut que tu te débarrasse du système de quelqu'un d’autre. Il
faut tout le temps que tu te poses la question : « Pourquoi elle gardait ça ?». C’est là que tu te rends compte
qu’elle était insécure, parce que des documents qui datent de 1997... Elle a tout gardé ! Tu ne veux pas trop
en supprimer, parce que tu te dis, est-ce que ça va me retomber dans la face un moment donné, peut-être
que je vais en avoir besoin comme exemple, etc. Alors il faut les prendre un par un puis regarder la date,
puis décider ce que tu conserves. Tu reprends les documents existants et tu les réorganises comme toi tu
penses. Je ne vois pas vraiment d'alternative à ça. ” (Manon)
121
structure produced by the secretary. Some of the main factors to consider are outlined
below.
89B199B5.4.1 Nature of the working relationship
The typical image of the personal assistant who answers the phone, manages the day
planner, types letters and organizes documents for his/her boss is just one possibility
among many. In some of the cases we encountered, the directors would have their own e-
mail inbox, manage their own documents, and answer their own phone. In another case
the secretary handled her boss’s documents, but not her boss’s e-mail. In fact, in some of
the cases we encountered, the support staff worker was more of a collaborator or a
colleague than a typical personal assistant. In short, there are many possible scenarios,
and every one of them will likely translate into a different document management
agreement between the two.
A more closely-integrated working relationship often translates into a single set of
documents for both, which in turn forces the secretary to directly consider the boss’s
needs at all times when making classificatory decisions. Because of that, the secretary
may not be able to organize things exactly as he/she would like to, but will instead adhere
to the guidelines established by the superior. The degree of freedom employees are
allowed in managing documents and the extent of their input in decisions related to the
documents structure are thus important factors to consider. Some superiors have
developed their own system and simply ask their secretary to use it as is, while others
build a system in collaboration with their employee:
R: Did your boss create this structure?
P: Well we did it together, I'd say. She said "Let's start like this", and after
that... it was a collaboration.F
66
66 R: Est-ce que c'est votre patronne qui a créé cette structure-là?
122
90B200B5.4.2 Time working together
We talked earlier about some of the coordination mechanisms that a secretary and his/her
superior may use (asking where to file a document, making classificatory suggestions,
etc.). These help ensure that the documents are placed where everyone expects them to be.
Such mechanisms are especially necessary in the beginning of the working relationship,
as the secretary doesn’t yet know the boss very well. At that point, he/she will be more
careful not to make too many assumptions when he/she is not sure where to store a file.
These assumptions could be based on the way things were with the previous superior, or
the way he/she would tend to do things herself.
What our participants reported, however, is that after a while, coordination mechanisms
become less necessary. With time, the employees come to know their superior and their
preferences, and thus learn how to better suit their informational needs.
When you arrive in a position like this, you start working with someone, and
you start to understand how she works. You have to work accordingly. You
come to a point where you know her, you know what she expects.F
67
By gaining a greater understanding of how the superior works, the worker is able to better
anticipate the superior’s needs and organize in consequence. Episodes of classification
uncertainty then become few and far between.
P: Ben c'est-à-dire on l’a fait ensemble je pense. Elle a dit « Bon, on va y aller comme ça », puis après ça…
Ça a été un travail commun… (Anne)
67 "Quand tu arrives sur un poste comme ça, tu commences à travailler avec quelqu'un, tu commences à
comprendre comment elle marche, elle. Il faut que tu marches aussi avec elle. Tu arrives à un point où tu la
connais, tu sais à quoi elle s'attend." (Carole)
123
This may eventually encourage workers to start taking more chances and simply classify
a document or create a new folder to the best of their knowledge. Making such guesses
does present a risk, as the employee may have to undo the changes later on if they’re
deemed unsatisfactory. But the risk of that happening becomes significantly lower after a
few years working with the same person. On the other hand, one worker told us she
would never create a folder on her own, seeing it as a lack of respect for the superior,
whose structure they were using.
91B201B5.4.3 Personality of the superior
Participants were quick to point out that not all bosses are the same. From one position or
replacement contract to the next, they come in contact with many different people, each
with their own methods, preferences and personality. One participant told us she felt
lucky because her current boss was a very organized person, who had developed a very
precise and efficient structure, so it had been easy for her to start working there and be
productive. In other cases, things may not be as idyllic. Sometimes, it is the boss that
relies on his/her assistant to help bring order to the chaos…
The superior’s personality can also have a direct effect on the secretary’s document
management behaviours, as the following anecdote illustrates:
You know, it also depends where you work. Before coming here, I replaced
someone in another department. Before she left, the girl I replaced told me
“You have to keep everything, keep a follow-up on everything, even the
delivery confirmations” – you know the kind of thing you receive that
confirms that the person has read your e-mail? She kept everything like that
because she had been scolded by her boss who was… not very nice, let’s just
leave it at that. Actually she was even a bit crazy I’d say. Because her boss
would often come back to check on her, she kept absolutely everything just to
be sure. She sort of transferred her fear to me, so when she left I kept on
124
doing that, and when I left I told my successor to do that as well. So basically,
if your boss is especially zealous, you’re going to keep everything. Not by
choice, but because you don’t want to get in trouble. With a boss who wants
to double-check everything you do, you will keep everything as evidence.
Whereas [my current boss] trusts me even though she doesn’t have any proof.
She just assumes things get done. She’s not going to watch me constantly to
make sure I did everything. […] So it really depends on the person with
whom you end up. Here it’s so different! I love it! F
68
Workers who move to a new position never quite know what to expect in terms of their
future boss. Although experienced employees with good connections will try to enquire
about the manager in place before making a move, not everyone has access to this kind of
information, or can even afford to pick and choose. One participant compared this to a
lottery, where you never quite know who you will end up with. And again, a period of
adaptation is usually necessary when starting a new working relationship:
“But it’s always like that. Everyone has their own style. But I’m a bit used to
this because in my 18 years in the private sector, and since I got here… [My
previous boss] was my 9th, and [my new boss] is my 10th. So I’ve changed
68 “Tu sais, ça dépend aussi où tu travailles. Avant de venir ici, j’ai remplacé quelqu’un dans un autre
département. Avant de partir, la fille que j’ai remplacé m’a dit “Il faut que tu gardes tout, garde un suivi de
tout, même les accusés de réception” – tu sais le genre of trucs que tu reçois et qui confirme que la
personne a bien lu ton message ? Elle gardait tout ça parce qu’elle s’était fait chicaner par sa boss, qui était
vraiment… pas gentille disons. En fait elle était même un peu folle, je dirais. Comme sa patronne revenait
tout le temps pour vérifier, elle gardait absolument tout pour être sûre. Elle m’a comme transmis sa peur,
alors quand elle est partie moi aussi j’ai fait ça, puis j’ai dit à la suivante de faire ça aussi. Bref, si ta boss
est zélée tu vas tout garder. Pas par choix, mais parce que tu ne veux pas avoir de trouble. Avec une boss
qui veut toujours revérifier si tu as fait ton travail, tu vas tout garder comme preuve. Tandis que [ma
patronne en ce moment] me fait confiance même si elle n’a pas de preuve de rien. Elle assume que ça s’est
fait. Elle ne va pas me regarder constamment pour être sûre que j’ai tout fait. […] Alors ça dépend vraiment
de la personne avec qui tu te retrouves. Ici c’est tellement différent! J’adore ça.” (Lise)
125
bosses quite a bit… Usually I give myself at least a month. Because you have
to understand how the person works. Like [my new boss] doesn’t work at all
like [his predecessor]. The other way was good too. And his way is OK too
but it’s just a different way. […] But it’s just a matter of communicating with
each other. You eventually adapt to the person, and afterwards everything is
fine.” F
69
However, there is obviously no guarantee that superior and secretary will be perfectly
suited to each other. In some cases, the two are simply not compatible for any number of
reasons. In such cases, the quality of the work may suffer and conflicts may emerge,
which can eventually lead to one of the parties leaving.
One participant insisted that liking the person you work for is a key factor in the quality
of your work, including document management:
You have to work for someone you like. You don’t work the same way for
someone you don’t like as for someone you like. Because if you like your
superior, you will want them to succeed. I see how this woman works, I work
with her, I work for her, and I feel involved. So I make sure everything works
as it should. If she asks me for some information, I want to be able to give it
to her quickly. […] That’s the difference; you have to have an interest. You
have to love your work. F
70
69 “Mais c’est toujours comme ça… Chacun a son style. Mais je suis un petit peu habituée à ça parce que
dans mes 18 années au privé, puis depuis que je suis ici… [Mon précédent patron] c’était mon 9ème, puis
[mon nouveau patron] c’est mon 10ème. Alors j’ai quand même changé pas mal de patrons…
Habituellement je me donne au moins un mois. Parce qu’il faut que tu comprennes comment la personne
travaille. Comme [mon nouveau patron] ne fonctionne pas du tout comme [le précédent]. L’autre façon
était bonne aussi. Puis lui sa façon est correcte mais c’est juste une autre façon. […] Mais il s’agit juste de
se parler. On finit par s’adapter à la personne puis après ça va bien.” (Édith)
70 “C’est qu’il faut que tu travailles pour quelqu’un que tu aimes. Tu ne travailles pas de la même façon
pour quelqu’un que tu n’aimes pas que quelqu’un que tu aimes. Parce que si tu aimes ton supérieur, tu vas
126
92B202B5.4.4 Change of superior
We have already discussed the high mobility of support staff, but in some cases, it’s the
superior who leaves, to be replaced by someone else. Although the worker stays in place
in this scenario, the effect of the change can be just as disruptive, as the new superior
may decide to change the way things are done. Not only can the existing classification be
modified, the worker may end up with much less (or much more) freedom than what they
were used to. Some of the methods or habits that they had developed when working with
their previous superior might now become useless, so it may be necessary to rethink or
abandon them.
One participant reported a situation in which a new superior arrived in a department. The
manager wanted to do a complete overhaul of the document organization, but didn’t want
to upset the secretary, who had been in this position for about 20 years. However, when
that secretary retired a few months later, the superior seized the occasion and decided to
start with a clean slate with the secretary’s replacement (our interviewee).
33B143B5.5 Worker characteristics
In a previous section, we downplayed the idea that only personal preference or
personality could account for how a worker handles documents, and instead highlighted
the influence of other factors in shaping a person’s document spaces. However, it would
be wrong to assume that individual characteristics have no impact at all. In fact,
individual characteristics appear to be one of the main drivers of variation among the
vouloir qu’elle réussisse. Moi je vois comment cette femme-là travaille, je travaille avec elle, je travaille
pour elle, puis je me sens concernée. Alors je m’arrange pour que tout fonctionne comme il faut. Si elle me
demande une information, je veux pouvoir lui donner rapidement. [...] C’est ça la différence ; il faut que tu
aies un intérêt. Il faut que tu aimes ton travail.” (Carole)
127
workers interviewed, in addition to these other factors (job status, relationship with the
superior, etc.).
93B203B5.5.1 Personality traits
Based on what participants told us, it appears that some personality traits may play a
significant role in explaining why they organize documents in a certain way. One of the
most often mentioned personality traits was flexibility or adaptability to new situations
(Pulakos et al., 1999). This is perhaps especially important in a high-mobility
organization such as the one we studied. When workers frequently change jobs, it is
obviously preferable if they can quickly adapt to the context in which they enter. Being
able to use the existing document structure is sometimes essential, as there are situations
where the worker is simply not at liberty to change it (superior’s structure, interim
replacement, etc.). Productivity in the early days of employment is thus largely
attributable to this capacity to adapt, and some appear to be better at it, or at least find it
easier, than others.
Another crucial aspect is risk tolerance. This has impacts on various aspects of document
management, perhaps most clearly in the propensity to keep documents or not (Jones,
2004). In our study, some workers were a bit more careful when weeding documents, and
tended to keep more than less, just in case. In the long run, this can of course result in an
unnecessary accumulation of documents, as we were able to observe in some cases,
where even the workers admitted that they had kept too much, but just couldn’t help it.
One participant attributed this to a question of self-confidence in your abilities and your
own judgment of the value of documents.
Interestingly, tolerance to risk and self-confidence appear to evolve over time. One
participant told us that she had become very careful about deleting documents after a bad
experience with a student. The two had an argument about a piece of information which
had initially only been written in an e-mail. However, our participant had deleted the e-
128
mail in question, and the student later came back to contradict the worker, who was now
unable to back up her assertion. The loss of this document had resulted in very unhappy
consequences for the worker, and had served as a lesson for the future. This suggests that
a single bad experience may be traumatic enough to permanently alter PIM behaviour.
Self-confidence may also affect other aspects of PIM, such as one’s willingness to
modify an existing document structure or start their own, or their propensity to make
suggestions to their superior about possible changes to their common structure. The level
of confidence is probably attributable at least in part to experience: more experienced
workers appear to be more likely to undertake an overhaul of the document structure, or
to discuss such issues with their superior on an equal basis. In return, the superior may be
more willing to accept suggestions from a more experienced, more accomplished
employee.
The idea of being organized or being an organized person also seemed very prevalent
among our participants. When explaining why their structure was especially detailed or
precise, some would explain that they were just that kind of person. The perception seems
to be that, by nature, some are just better at organizing and classifying things, and that
this extends to any field of activity, including their personal life. Some of our participants
recounted anecdotes about their early school years, and how their desk and their day
planner were carefully organized, and that such habits had continued to this day, and now
transpired in the organization of their work documents.
I’ve always been like this, I’ve always been super-organized. […] I like
organizing things. Maybe sometimes I’m a little annoying but… I’m not a
perfectionist, but I like when things are in their place. Of course, it’s subject
to interpretation but I think I have pretty good judgment, which means that I
can make sure that people will find their way around, generally speaking. F
71
71 J’ai tout le temps été comme ça, j’ai tout le temps été super-organisée. […] Moi l’organisation j’aime
bien ça. Je suis peut-être des fois un petit peu fatigante, mais… Je ne suis pas une perfectionniste mais
j’aime que les choses soient à leur place. C’est sûr que c’est de l’interprétation mais je pense que j’ai quand
129
This quote brings forth the idea of not only being good at organizing documents, but
actually enjoying it. Five of our participants said they mostly enjoyed that aspect of their
job, and generally didn’t view it as a burden, even though at times things are a bit less
rosy, for example when arriving in a position where things have been left in a rather
messy state.
Personality is a complex topic, and a thorough assessment of our participants’ personality
(using standardized personality tests, for example) was initially considered beyond the
scope of our exploratory study. In retrospect however, given the apparent importance of
this aspect within our participants’ responses, it might have been useful to systematically
administer such tests to our participants before or after interviews, and then try and make
links between specific PIM-related behaviours and particular personality traits as
revealed by the tests. For example, did the participants who exhibit similar attitudes on a
particular aspect of document management also share similar personality traits? It is our
opinion that this would make for an interesting follow-up study.
94B204B5.5.2 Preferences and dislikes
Although closely linked to personality traits, some of the opinions put forth by our
participants appeared to be due to personal preferences or pet peeves. They frequently
expressed their preference or dislike for a particular aspect of PIM, sometimes using very
expressive vocabulary (“I hate that”; “I just can’t stand that”; “It really gets on my
nerves”, etc.).
Some examples include keeping documents on the Windows desktop, having more or
less folders at each level of the hierarchy, using paper documents, etc. There was
même un jugement qui fait que je suis capable de faire en sorte que les gens vont se retrouver, en général.
(Lise)
130
sometimes a logical explanation for these preferences, but at times, participants had
trouble explaining exactly why something would “get on their nerves”. It just did.
95B205B5.5.3 Training and education
Most of the workers we interviewed had received some sort of formal training in
document organization, whether during their Cegep years or during another secretary
course, or through the university’s Archives department. Some participants reported
learning some useful advice from these training programs, even though what they learned
might not be directly applicable to their current situation, for a variety of reasons. Such
courses can at least expose students to some proven methods, which they may be able to
apply at least in part over the course of their career.
Although it’s hard to determine exactly to what extent, training has the potential to make
a difference in a worker’s career, especially in the beginning. After a while though, it
appears that experience is more significant in helping workers perform better in
document management tasks. Still, professional training can help workers get a good
head-start in their career, instead of just learning everything through the trial and error
that comes from experience.
96B206B5.5.4 Experience
Beyond formal education, most participants seemed to agree that experience was a key
factor in PIM. Through experience, workers are exposed to many different situations, and
accumulate various tricks and techniques, either by learning from others or from their
own experiments, successes and mistakes. High mobility within the organization also
means that a worker can potentially go through several different departments and
contexts, and get the chance to work with many superiors, each with his/her own
idiosyncrasies. They can thereby accumulate a wealth of practical knowledge that is
typically not available to new workers.
131
Having experience has several potential advantages. For one thing, it allows you not to
repeat the same mistakes you may have committed in the past. It can also shorten the
adaptation period required at the start of a new position, or simply when new situations
arise.
From our interviews, we were able to identify three types of experience that may prove
useful: experience in a similar position (e.g. office technician, administrative assistant,
secretary, etc.), experience in a similar organization (e.g. the academic environment) and
finally experience within the current organization.
In the first case, the person will have a good overview of what working in an office – and
organizing documents in general – involves. The person also has a good understanding of
what his/her role is, and can usually anticipate the needs more easily. They may have
useful insights into how documents should be organized to maximize efficiency.
However, the type of organization in which one has worked also matters. There are
probably positions of office support staff in most organizations, but some are quite
different from one another. One of our participants had a lot of experience working in
notaries’ offices, and claimed that she had become very proficient at maintaining her
documents, and those of her bosses, in good order. She explained the importance of
experience in managing documents:
A girl who’s just getting out of school, who has no experience in notarial
service and who arrives in a notary’s office and tries to create a structure, she
won’t know how, because she doesn’t know who are the stakeholders related
to a file. But I was experienced, I had worked in the industry for 25 years, I
understood all of that, so I knew there would be banks, surveyors, cities and
all sorts of people involved, and the role of each one. When we started a
project, for example a new construction project, I knew about all the
documents, where they came from and to whom they were going, and I was
132
able to structure everything very well. First within a folder, and then in the
file cabinet. […] I learned everything on the job. But after 25 years I had
accumulated so much experience in the industry that in the last few years,
when I arrived in a new office, they gave me the go-ahead to organize
everything the way I wanted. I had experience and I had a vision of the work
I had to do. I understood the work of the notary, and I understood my part. F
72
Interestingly however, when this participant arrived in the academic environment, she
seemed to lose her beacons, and when we met her, she was still struggling to make sense
of her new environment even after over a year in her new position. She was in a situation
where after spending her entire career in one industry, she had moved to a different one,
and thus needed to create a new mental model of her surroundings. In a way, it is as if her
former expertise was almost too entrenched, and she had trouble putting it aside to adapt
to a new context.
Finally, experience within the current organization is also considered very precious, as it
helps give workers a sense of what their place in the organization is, and where people
and documents fit in the greater scheme of things. Such workers may already know about
the various committees, instances and acronyms used in the organization, which will
facilitate their integration in their new department.
72 “Une fille qui sort de l’école, qui n’a pas d’expérience en notariat puis qui arrive dans un bureau pour
faire sa structure, elle ne saura pas, parce qu’elle ne sait pas quels sont tous les intervenants dans un dossier.
Mais moi j’étais une fille d’expérience, j’avais travaillé dans le domaine pendant 25 ans, je connaissais tout
ça, alors je savais qu’il y avait les banques, les arpenteurs, les villes et toutes sortes de monde, puis le rôle
de chacun. Quand on partait un projet, par exemple un nouveau projet de construction, je connaissais tous
les documents, d’où ça vient et à qui ça va aller, et je pouvais tout bien structurer. D’abord à l’intérieur
d’un dossier, puis dans un classeur. […] J’ai appris sur le tas. Mais après 25 ans j’avais tellement
d’expérience dans le domaine, que dans les dernières années, quand je suis arrivée dans un nouveau bureau,
on m’a donné le feu vert pour tout organiser à ma manière. J’avais de l’expérience puis j’avais une vision
du travail que j’avais à faire. Je connaissais le travail du notaire, et je connaissais ma partie à moi.” (Kim)
133
97B207B5.5.5 Computer literacy
Computer literacy can be defined as “the basic knowledge, skills, and attitudes needed by
all citizens to be able to deal with computer technology in their daily life” (Tsai, 2002, p.
69). In the workplace, it usually means “the ability to use specific software applications
for well-defined tasks, such as word processing, e-mail, spreadsheets, and Internet
searches” (Bers, 2010, p. 14), which roughly corresponds to what was asked of the
workers we interviewed.
None of our participants considered themselves computer-illiterate, and in fact quite a
few felt that they were experts at using software in the context of their work. The
youngest participants, in their twenties and early thirties, usually seemed to take
computers as a given, while some of the older participants had started their career in the
world of typewriters and had to adapt to newer technology as it surfaced. Nevertheless,
most had come to embrace the use of computers, and were able to perform at least basic
document management tasks in the Windows environment.
There is evidence in the literature that older workers tend to have lower computer literacy
than their younger counterparts (Chiu, 2000; Lin, Lin, Jiang & Lee, 2007). That being
said, in our study, some of the older workers considered themselves quite proficient, and
with good reason. They had often made efforts to keep abreast of newer software, notably
by regularly attending continuing education classes on the various modules of the
Microsoft Office software, the Web, and other relevant topics. We were able to witness
the ease with which they navigated in the Windows environment and used Microsoft
Office applications.
It remains unclear to what extent computer literacy could be a significant factor in
helping a worker organize his/her documents. That being said, a person who is more
comfortable with computers in general may be less reluctant to rely on and use electronic
134
documents, instead of prioritizing the paper format. We did notice that in general the
younger participants in our study seemed much less attached to the physical version of
documents than some of their older colleagues, some of whom told us they much
preferred keeping both versions:
It’s because they’ve been doing things that way for a while, and [my boss] is
comfortable that way, so… But me, I don’t think I’d print everything. I’d put
everything in digital folders, but I wouldn’t make paper folders in addition to
that, but [my boss] likes it that way. […] Personally, I prefer using the
electronic documents than paper anyway…F
73
However, even some of the older participants, who had always favored paper documents
and taken them for granted, were beginning to question their old ways. We asked one
participant, who maintained both digital and paper files, where she would turn first if she
was looking for a document: the digital files or the paper files? After hesitating a bit, she
admitted she would probably turn to the computer first, which seemed to surprise herself
the most:
It’s interesting that you’re asking me this question. That’s where I’d go:
digital. Not bad, eh? ... I’ve just had an existential epiphany! [laughs] In the
end, you make me realize that I’ve finally accepted the digital. F
74
73 “C’est parce que ça fait quand même longtemps qu’ils fonctionnent comme ça, et [ma patronne] est à
l’aise comme ça, alors… Mais moi il me semble que je n’imprimerais pas tout. Je mettrais tout dans les
dossiers électroniques mais je ne me ferais pas des dossiers papiers en plus, mais [ma patronne] aime ça
fonctionner comme ça. […] Personnellement, j’aime mieux consulter les trucs électroniquement que papier
de toute façon…” (Olivia)
74 “C’est drôle que tu me poses cette question-là. J’irais là : électronique. Quand même pas pire… Je viens
faire une prise de conscience existentielle. [rires] Finalement tu me fais prendre conscience que j’ai
finalement accepté l’électronique.” (Anne)
135
Our question seemed to prompt her own realization of the extent to which digital
documents had become a significant component in her work, but also the extent to which
she had become comfortable with them. So although it is still quite prevalent in today’s
work environment, the importance of paper may diminish in the coming years, not only
due to the arrival of a generation of younger workers, but also as more of the veteran
workers realize that they’ve “accepted the digital”.
98B208B5.5.6 Cognitive style and abilities
There is evidence within the literature that an individual’s cognitive style or abilities has
an influence on PIM performance. Aspects such as flexibility of closure (Gwidzka, 2004),
working and visual memory (Gwidzka & Chignell, 2004), as well as spatial ability
(Modjeska & Chignell, 2003) have all been highlighted as possible factors in previous
studies. Aside from the few studies mentioned above, however, this remains a relatively
unexplored topic within PIM.
As was the case for personality tests, cognitive abilities were not formally measured or
evaluated in our exploratory study. Yet they do deserve a mention here, because they may
very well be playing a role in the behaviour exhibited by our participants, and therefore
may account for some of the variation observed. For example, a participant mentioning
that they weren’t able to function a certain way and “had to” do things another could be
attributable to specific cognitive traits that they may or may not have. Again, this could
certainly be the focus of a follow-up study.
34B144B5.6 Organizational context
Employees do not work in a vacuum, so it is also necessary to consider the organizational
context in which they evolve in order to fully understand what they do. Perhaps
somewhat indirectly, an organization’s mission, core values and culture have an impact
on the way its employees will work. These elements can affect workers’ priorities, and
136
thus the way they manage their documents (e.g. what should be kept). Such impacts,
however, can be hard to measure.
On the other hand, organizations can also establish more concrete policies, rules or
guidelines, which may affect PIM behaviour more directly, by modulating, constraining
or even replacing whatever system the workers may have created on their own.
We were able to observe the effects of such a top-down initiative in our study, in the form
of the university’s organization-wide institutional classification scheme. The scheme was
developed by the Archives Division of the university, although its use is not strictly
enforced. It can be implemented on a voluntary basis by any service or department within
the university, and some guidance from the Archives Division is usually available during
implementation.
Only three participants in our study were using the system, although one was using it
somewhat unknowingly; she was a recently-arrived employee and had not realized, nor
been told, that her predecessor was using the official system.
Although few of our participants were actually using the university scheme, several
others had used it in the past, had heard about it from colleagues who used it, or had
received the required training from the Archives Division, even if in the end they had not
implemented it. Two participants did tell us that even though they had not implemented
the system, they were able to take away some very good general advice about document
management from this training.
When asked why they had not implemented the official system, some spoke of the
classification’s complexity, but most of all that it was not always easily applicable to
local needs. The idea that it was too broad or general for the needs of a single, usually
more specialized, unit was a common complaint. We sensed that the institutional scheme
was more commonly-used by (and perhaps better-suited to) higher management and their
assistants, because they had to oversee a broader spectrum of activities, and thus could
137
make use of the entire scheme. In that case, such a broad classification makes sense. It
may however be less interesting for middle managers, who tend to focus on a narrower
domain of activities. If one’s entire scope of action fits within only one or two main
classes in a classification system, they may not view it as very relevant or useful to their
needs. In that case, a locally or individually-developed system may appear preferable.
Having to learn a new classification system, and thus potentially abandoning your own or
at least the one you had been using for a while is also a daunting prospect. However, such
a transition is certainly not unimaginable for these workers, who are used to changing
jobs – and classification systems – regularly. Once again, the key word here is adaptation.
But perhaps the most discouraging aspect, as reported by our participants, is the actual
conversion process from one scheme to another, which means examining every existing
document and reclassifying it according to the new guidelines. In some units, where there
was a few decades’ worth of documents, this was a rather intimidating, if not entirely
discouraging prospect. This type of task is very time-consuming, even with some outside
help, and since document organization is not really part of the core mission of most
services and departments, it is likely to take a backseat to more pressing concerns. As one
participant put it, implementing this system is an ideal, but not an easily-attainable one.
The system was perhaps described as an ideal because it does present several concrete
advantages. It gives everyone a common basis to organize documents, which will remain
the same as employees come and go. In some instances, everyone in the unit is
encouraged or asked to use the system, which makes for a cohesive document
environment within the unit. The classification system also comes with a guide that gives
workers indications on how to classify some of the more ambiguous documents. It is
unlikely that any worker-developed classification could ever be as well-documented. An
additional advantage is that the system helps units plan and enact efficient and regular
weeding of their documents once they become inactive. This ensures that documents do
not accumulate for years, as we were able to witness in several instances. Software has
138
even been developed to automatically alert employees when it is time to prepare boxes of
documents to be archived.
The participant we interviewed who did go through this entire conversion process said
that it was indeed a good opportunity to make a significant and much-needed cleanup and
weeding operation, and that she was satisfied with the new system - and the newfound
space in her file cabinets! She had struggled at times with the classification, and
sometimes still had to convene with her superior to decide how to use it to suit their
particular, local needs. The system does include a provision for adding local sub-classes
within certain pre-defined classes, and under certain conditions.
However, our participant also emphasized that the transition had been a long and difficult
journey, and that she wouldn’t have succeeded without the help of the Archives employee
who visited her regularly during the transition. And unfortunately, she told us that she
had received special help from the Archives Division because she made the transition as
part of a pilot project, but that other units in the future might not be able to get as much
assistance.
In fact, despite very interesting advantages, most workers interviewed (and probably their
superiors as well, who were likely to make the final decision in such matters) felt that the
process was just too demanding to actually go through with it. In other words, the barriers
to entry remained more important than the expected positive outcomes. In the end, it
seemed that only the most courageous, or perhaps the most desperate, were willing to
undertake such an endeavor.
35B145B5.7 Document attributes
We have discussed contextual factors, factors related to the employee and the content of
their job, but we also need to examine the nature of individual documents. Within a
worker’s document space, not all documents are necessarily of equal value. Some will be
considered more useful or important, and will thus be handled differently. More attention
139
may be paid to that document’s location (put on the desktop for easy access, use of
desktop shortcuts) or filename (use of special characters to drive the file to the top of the
list), or the document may be more frequently backed up to avoid any risk of losing it.
In our pre-interview questionnaire, we asked participants to identify some of the
documents they considered more important within their document space. During the
interview, we then asked them to explain why they had chosen these particular
documents. We could then observe whether or not special treatment had been given to
these documents.
Generally speaking, documents can be considered important if they are central to a
worker’s task. But beyond this, we were able to distinguish three main characteristics of
important documents.
First, workers will attach more importance to documents that relate directly to one of
their main responsibilities, especially if they are the sole person who is responsible for
that particular task:
R: So “Maintenance contracts.xls” is important to you?
P: Oh yes. That’s because I’m the one who takes care of this. It is really,
really, really [sic] important for me because it’s my main reference.
Because this file contains all of our contracts’ expiry dates, so when a date
approaches, I absolutely have to know it in order to jumpstart the renewal
process. If I miss that… the entire university has a problem! For sure I often
access that file, and I even put a shortcut on my desktop because I use it
pretty much every day. F
75
75 “R : Alors “Contrats d’entretien.xls”, c’est un document important pour toi ?
P : Ah oui. C’est parce que c’est moi qui s’en occupe. Ça c’est vraiment, vraiment, vraiment [sic] important
pour moi parce que c’est ma référence. Parce que ce fichier-là contient toutes nos dates d’échéances de
contrats, alors quand la date approche, il faut absolument que je le sache pour pourvoir enclencher le
140
This worker relies on that file in order to accomplish perhaps the most crucial aspect of
her job, the one for which she is held responsible within her unit. This warrants a sort of
special status for the document, and a place on the desktop.
This quote also brings us to the second characteristic of important documents, which is
frequency of use. If a document is used on a regular basis, for instance on a daily basis, it
might be worth putting it in an easily accessible location within the document structure.
In the long run, not having to navigate too much to access the document can save a lot of
time. Again, the desktop is a popular location for that purpose, but we have also seen
documents left intentionally as orphan files within the main folder, as this gave quicker
access than if the file had been buried a bit deeper within a sub-folder.
The third characteristic of important documents is that they are especially useful in
current matters. A file may have been very important in the past, but if it relates to a
project that was completed years ago, it might not retain much value today. Important
files are usually those that relate to current tasks and projects. That being said, some of
the files deemed important by our participants were not necessarily related to very current
concerns. In fact, we identified two additional criteria that might lead a worker to attach
specific importance to a document.
The first is the amount of effort that went into the creation of the document. If a worker
has invested a lot of time and effort, they will be weary of losing the document in
question. There is of course the fear of having to do the work all over again, but there is
also the idea that the file represents the final product of a lot of invested effort, even if
that final product is not really used regularly anymore. That document does, then, stand
as a sort of achievement or trophy, and is thus something that they would hate to lose:
renouvellement. Si je manque ça… c’est toute l’université qui a un problème ! C’est sûr que je vais souvent
dans ce fichier-là, puis d’ailleurs j’ai un raccourci parce que je m’en sers tous les jours ou à peu près.”
(Ingrid)
141
R: When I asked you to identify documents that you consider especially
important to you, the first one you mentioned was the "Activity Report".
P: Well, that's what I've been working on since I've arrived here actually. It's
the most important thing. You see it's about 30-40 pages long. So it's
something I definitely wouldn't want to lose. I think I'd cry! F
76
Models or templates workers may have developed were also mentioned as things they
treasured, as these could be helpful even in future positions, and again, a significant
amount of work had gone into their creation, in order to save time later.
The second aspect is exclusivity. Workers know that some of the files they have are
somewhat exclusive to them, in that they are probably the only person to have a copy.
This scarcity obviously gives those documents additional importance in their eye, which
wouldn’t be the case if such a file were available to all in the local intranet. Some told us
that they couldn’t really trust their superior to keep a copy, so they made sure they had it
on hand. They knew that at any moment they could be asked by someone to provide that
document (their superior, but also colleagues or even external clients). They thus acted as
a kind of custodian for these documents, and obviously took great care of them.
The special status allotted to certain documents is thus one more of the factors that help
make every worker’s document space somewhat unique.
76 R: Quand je vous ai demandé d'identifier les documents que vous considérez particulièrement importants,
le premier que vous avez mentionné c'était le "Rapport d'activités".
P: Bien, en fait c'est pas mal là-dessus que j'ai travaillé depuis que je suis ici. C'est la chose la plus
importante. Tu vois, c'est à peu près 30 ou 40 pages. Alors c'est quelque chose que je ne voudrais pas
perdre, c'est clair. Je pense que j'en pleurerais ! [Ginette]
142
36B146B5.8 Development of a conceptual model
Having so far identified the various factors that are likely to influence PIM among office
support staff workers, we can now gather these elements within a single model, in order
to obtain a more complete view of the influences at play. Figure 5.1 presents an overview
of the main types of factors we have identified in this chapter.
It is interesting to note that although entirely grounded in the data gathered from our
participants, our model contains (albeit in a slightly different configuration) all of the
factors identified in our initial conceptual framework (see figure 2.2, in section 2.6).
Perhaps even more interesting is that several of the factors we have identified also figure
in Sabine Mas’ (2007, p. LXXXIX) conceptual framework for her study of retrieval
efficiency in personal classification schemes. Though she admitted that the influence of
the stated factors had not been empirically verified in her study, her conceptual
framework, which was derived from previous literature, included elements such
organizational context (institutional policies in Mas’ study), worker characteristics
(training, experience), job content (professional role) and document attributes (value), all
of which also figure in our model.
143
Figure 5.1: A conceptual model for personal information management of office
support staff workers.
Our model shows that when engaging in any type of PIM behavior, workers are likely to
be influenced by elements of one or more of these seven categories. This, in turn, will
have an influence on the characteristics of the resulting document spaces. For example,
when deciding whether a document should be weeded, a worker is likely to be influenced
by several factors, such as the perceived value of the document (document attributes), the
worker’s risk tolerance (worker characteristics) or whether or not they have the freedom
to do so (relationship with the superior).
The model presented here thus forms a conceptual approach to of our research, and
highlights the multiplicity and the complexity of the factors involved in shaping a
worker’s document spaces, a fact that should be taken into consideration by every
researcher in the PIM field.
PIM Behaviour
4)Relationship with the superior
5)Worker
characterisics
6)Organizational
context
7)Document attributes
1) Job Content
2) Existing
documents
3) Job status
Document
spaces
144
Therefore, when observing a worker’s documents, one should be aware that there might
be several reasons why a certain item ended up in a specific location, or why a worker
has developed particular behaviours. One should acknowledge every component of the
model as a potential factor, and should consider the potential interactions between these
factors. This could lead to a better, more in-depth understanding of the way workers
organize their documents.
37B147B5.9 Limitations of the study
As we mentioned earlier, this study was relatively limited in its scope, in terms of the
number of participants interviewed (15) and in the number of organizations surveyed
(one), both of which met the requirements for the application of grounded theory. The
results may be different in a smaller organization, or in an organization using a different
document management software environment. There was also uniformity in the gender of
the participants, although this was an almost unavoidable consequence due to the makeup
of the targeted positions. Therefore, we cannot make strong claims to the extensive
generalizability of our data.
However, this was not the goal we initially pursued. This study was purposely
exploratory in nature, and merely intended to provide a first glimpse into relatively
uncharted territory. Furthermore, grounded theories are substantive theories, which do
not strive for universal generalizability. We must therefore consider the strong possibility
that additional factors could eventually be added to our model. The conceptual
foundations we have established should be tested and perhaps modified in light of new
discoveries. Our findings would thus undoubtedly benefit from any similar studies
conducted in the future.
Despite the limitations identified here, this study has allowed us to discover many
interesting elements with regard to PIM in the workplace, several of which had not been
mentioned in the existing literature. In other cases, we were also able to confirm findings
145
from other researchers in the field. As such, this study has more than reached its
exploratory objectives.
146
7B117BChapter 6: Conclusion
38B148B 6.1 Summary of the findings
The initial research questions for our study were as follows:
How do office support staff organize their documents?
Why do office support staff organize their documents the way they do?
Chapter 4 provided answers to the first question. In that chapter, we identified six main
components of workers' document environments, which we called documents spaces: the
main folder, the secondary folders, the Windows Desktop, e-mail, shared environments
and paper documents. Most participants used all of these, though not necessarily to the
same extent or in the same manner.
Along the way, we also highlighted workers' specific behaviours pertaining to elements
such as the handling of orphan files, the management of multiple versions, the naming of
files and folders, searching and browsing among files and weeding. And again, while we
were able to identify some common ground among participants, there was also significant
variation in the way things were done.
As for the question of why workers organize their documents the way they do, we
presented in chapter 5 a series of factors which are likely to influence document
organization. The factors were grouped into seven major categories, which we summarize
below, along with significant examples:
1. Job Content: The breadth or complexity of a position can influence the number of
first-level folders, which can be seen as a validation of Malone’s (1983) findings.
Furthermore, a higher work pace may have negative effects in the attention given
to organizing documents, and may encourage the placement of frequently-used
147
documents in easy-to-reach places. The effects of work pace on PIM performance
constitute an original finding, of which we saw no previous mention in the
literature.
2. Job Status: Workers tend to behave differently whether they are in a permanent
position, or if they merely act as temporary replacement, which can have an
impact in the care given to the organization of documents, notably by
discouraging long-term thinking and the undertaking of major document structure
overhauls. As far as we know, the impact of job status on PIM has not been
considered in past studies, yet appears to have significant consequences on the
way workers effectively manage information on a daily basis.
3. Existing documents: The state of the existing documents related to a position can
either help workers acclimate to their new environment, or else be a significant
hindrance to their initial and ongoing efficiency at work. Not all workers receive
assistance from their predecessor, yet they ultimately have to decide on an
appropriation strategy, such as continuing with the same organization or starting
anew, and carrying over the old files or not. This transitional aspect in PIM has
also been neglected by previous studies, where it often seems to be assumed that
workers are the owners of their files, know exactly what files they have under
their control, and are working within a structure that they have developed. Our
research shows that in reality, the situation is often very different: support
workers regularly have to deal with a structure that has been at least partially
developed by someone else, and they may possess files they don’t even know they
have.
4. Relationship with the superior: Workers in a secretarial or personal assistant
position do not work in a vacuum. They may have to manage someone else’s files
in addition to theirs, or have files in common with that person. Elements such as
the length of time working together and the personalities of both parties can
influence workers’ PIM behaviour. Additionally, a change in superior may have
148
significant consequences on a worker’s document organization, as new rules,
methods and customs may come into effect. Therefore, office support staff often
have a partial or slightly constrained freedom in organizing their documents. This
aspect has not really been touched upon in PIM studies, where participants
generally appear to enjoy complete freedom with regard to their documents.
5. Worker characteristics: Individual differences such as personality traits,
preferences and dislikes, training and level of computer literacy can all affect
workers’ document organization. As in other domains, experience is also a major
element, as more seasoned employees learn from previous mistakes and usually
possess deeper contextual knowledge of the organization and its specificities. All
of this makes generalizations very difficult when trying to determine how workers
normally handle documents. In fact, one may be tempted to say that no two
workers are exactly alike when performing PIM tasks, which obviously has
important implications for researchers and software developers (Gwizdka &
Chignell, 2007).
6. Organizational context: Every organization is somewhat different, and most
importantly every unit within an organization may have its own peculiarities,
traditions and culture. Not all of these elements consistently affect PIM, but the
use of an institutional document classification scheme or other departmental rules
may have a significant impact on the organization of documents even at the
individual level. This relates to the contextual component of PIM, as put forth by
Kwasnik (1991), Barreau (1995) and others.
7. Individual document attributes: Beyond general habits and preferences, not all
documents within a corpus will be treated the same way. Aspects such as the topic
of a document are obviously considered when deciding on its location, but a
document’s frequency of use, perceived value or relevance to current concerns
may also translate into special placement within the document structure. This
once again confirms Kwasnik’s (1991) assertion that document and situational
149
attributes can both have an effect in document organization decisions. It also
reinforces the notion that importance can be a significant factor in PIM decisions
(Bondarenko & Janssen, 2005; Whittaker & Hirschberg, 2001; Paré, 2007).
Using quotes from participants along with our own observations of their document
organization, we were able to demonstrate how the above factors can directly influence
and ultimately shape a worker's document spaces. The conclusions and contributions of
our study are thus undoubtedly grounded in the reality of the participants.
39B149B6.2 Significance of this study
We believe that this study constitutes a useful contribution to the area of PIM and its
related disciplines, for several reasons.
First, it is one of the few PIM studies to focus on office support staff. This type of
workers had been somewhat neglected in previous research, but our findings confirm that
they face several unique challenges regarding document management. The need to adapt
to a superior's document organization, as well as the frequent employee turnover
observed in this type of position present particular PIM-related challenges, which should
be taken into account when developing document management systems and policies. That
being said, a lot of what we have observed could potentially apply to other types of
workers, so in that sense our research should provide interesting insights beyond those
concerned with office support staff.
Secondly, our study highlights the complexity involved in shaping a person's document
environment and document management practices. While it may be tempting for
researchers to focus on one or a few determining factors, or to pigeonhole people into
prototypical categories or profiles, our results show that a very large number of factors
have to be taken into account when studying a person's PIM behaviour. A particular way
of doing things may be attributable in part to personal preference, but also to the nature of
the position, or even the state of the predecessor's document organization, among other
150
factors. This should serve as a warning for future researchers to conduct macro-level
analysis and adopt a more holistic approach in the study of PIM. The same goes for
developers of tools such as document management software, e-mail applications and
operating systems.
Thirdly, this study contributes a new conceptual model of the factors that may influence
document organization by office support staff. Because it is the product of a single study
conducted within a single organization, we do not expect it to be a universal
representation of reality in the workplace. That being said, it is our hope that future
researchers in the field can use it as a basis for their own work, and in turn help further its
development from a substantive theory to one that would be more widely applicable
(Glaser & Strauss, 1967).
Again, it should be noted that some of the factors we identified in our model were, as far
as we know, not part of the existing literature on PIM. Job Status, Existing Documents
and Relationship with the Superior were all identified as crucial elements, yet had not
been mentioned by previous studies. These elements are especially interesting in that they
highlight potential PIM problems caused by employee mobility in organizations. This
obviously has serious consequences on knowledge and continuity management in
organizations (Beazley et al., 2002).
Furthermore, this study used a grounded theory approach to data collection and analysis.
This is not uncommon in the field of library and information studies (Seldén, 2005;
Mansourian, 2006), but much rarer within PIM. The present study will hopefully serve to
show that grounded theory is indeed a viable methodology in this field of study, as it
allows us to derive theoretical constructs from the reality observed in the field.
Finally, in addition to some of the original contributions of our study, we were able to
confirm findings from previous PIM studies, thereby giving added weight and credibility
to some of the major contributions to the discipline. Here are some of the findings for
which we were able to add corroborating evidence:
151
Workers’ sense of guilt with regard to their document organization (or lack
thereof) (Jones et al., 2005; Boardman & Sasse, 2004).
The lack of time workers usually devote to assessing their document organization
practices (Jones (2007).
The use of the Windows desktop, notably for shortcuts to the most used software
applications (Ravasio, Schär & Kruger, 2004)
The importance of the reminding function in both virtual (Barreau & Nardi, 1995)
and paper (Malone, 1983) environments.
The lack of consistency in structure between a person's various document spaces,
e.g. e-mail folder structure vs. Windows folder structure (Boardman et al., 2003).
Several elements from the conceptual framework proposed by Mas in her 2007
study of retrieval efficiency in personal classification schemes.
40B150B6.3 Future research and recommendations for researchers
In terms of future research, several interesting avenues could be explored. In fact, every
single factor included in our model could benefit from more focused attention. That alone
could make for several substantial studies. For example:
The role of personality could be further examined, perhaps through the use of
personality tests.
In the same vein, cognitive tests could help reveal the links between cognitive
abilities, such as memory and spatial ability, and PIM performance.
152
Our understanding of the relationship with the superior could probably be
improved by interviewing both the worker and his/her superior. In our study, we
focused on the worker's viewpoint, but interviewing both parties might generate
additional insights.
However, as we have stated earlier, any results from such focused studies would have to
be considered within the broader context, in combination with the other factors included
in our model. This is perhaps the most crucial lesson researchers can take from our
findings.
It would also be useful to conduct similar studies in other organizations of various sizes,
in different industries, in both the public and private sector, with and without a worker's
union, etc.
The impact of the implementation of document management software such as Microsoft
SharePoint is also an interesting avenue, because such an implementation is likely to
induce significant changes in PIM behaviour. A before-after study, or perhaps a
comparative study between organizations that use such software and others that don’t,
could yield interesting data.
Finally, it would be interesting to see to what extent the model we have developed applies
to other types of employees beyond office support staff (interns, middle managers, etc.).
It is conceivable that a large portion of the factors we have put forth would also apply in
such cases, but we would also expect some differences.
In other words, there is significant potential to build on the present study, and to further
contribute to the theoretical foundations of PIM.
41B151B6.4 Recommendations for practitioners
153
Given that our study was firmly grounded in the daily reality of our participants, it seems
appropriate to also include a few concrete recommendations for practitioners working in
the field (archivists, knowledge managers, managers, etc.). Although PIM is done on an
individual basis, it is apparent that organizations could do more to help facilitate this task.
This responsibility could very well be given to specialists in information studies and
knowledge management.
Throughout this study, high employee turnover was frequently mentioned as a source of
problems in terms of its effects on document management. Dealing with a predecessor's
existing documents and building a document structure for a new, unfamiliar position
were seen as significant challenges by our participants. In light of these concerns, we
hereby formulate two recommendations to organizations regarding document
management.
The first involves the training of new employees. These employees are often trained in
the basic tasks and requirements of the job, but rarely receive much advice regarding the
current organization of documents. This can potentially make for a more difficult arrival
in their new position, as they have to either guess how their predecessor functioned, or
start a new structure from scratch. It can also translate into a significant amount of time
and effort being wasted. Therefore, special attention should be given to the transition
between a predecessor and his/her successor. In the best cases, when the predecessor still
works within the organization, they should be made available to provide some basic
training on the current document organization. In fact, closer attention should be given to
training the employees not just in how the work should be done (i.e. in terms of
procedures and daily operations), but also about the context in which they are going to
work, especially with regard to some of the specialized vocabulary used. This would
probably be worth the time investment, as it could significantly enhance workers’
productivity from the very start of their employment.
Alternatively, it would be a good idea for workers to provide an outline of the structure
they have developed, along with explanations of the types of documents that can be
154
found in each folder, perhaps even a sort of index. As we stated earlier, the logic behind a
person's document structure can be considered tacit knowledge, so making it explicit by
putting it to paper should be helpful for future replacements, and current colleagues as
well. This would facilitate the transition between employees.
In a similar vein, our second recommendation is for organizations to seriously consider
the establishment of an institutional or perhaps a unit- or department-level scheme for the
organization of documents. While a standard scheme certainly presents some drawbacks,
notably in terms of individual freedom and flexibility, it also has significant advantages.
Standard systems are usually well documented, which should help new employees
understand it more easily. Additionally, if everyone in a department uses the same system,
any employee can potentially train newcomers and provide them with punctual help, as
needed. Although some level of effort has to be expensed in developing it, such a system
is likely to help maintain the continuity of operations within an organization.
Most importantly, we wish to reiterate the importance of a macro-level investigation
when analyzing the PIM behaviours and processes within an organization. The model we
have established undoubtedly demonstrates that PIM is influenced by numerous factors,
which all need to be considered before taking major systemic and organization-wide
decisions.
155
8B118BBibliography
Abrams, D., Baecker, R., & Chignell, M. (1998). Information archiving with bookmarks:
personal web space construction and organization. Proceedings of the SIGCHI
conference on Human factors in computing systems, Los Angeles, California, 40-48.
Agarawala, A. & Balakrishnan, R. (2006). Keepin’ It Real: Pushing the Desktop
Metaphor with Physics, Piles and the Pen. CHI 2006, April 22-27, 2006, Montréal,
Québec, Canada. Hhttp://honeybrown.ca/Pubs/BumpTop.pdfH.
Alavi, M. & Leidner, D. (1999). Knowledge management systems: issues, challenges,
and benefits. Communications of the AIS, 1(2), 2-36.
Alkhalifa, E. M. (2006). Cognitively Informed Multimedia Interface Design. In: Ghaoui,
C. (Ed.). Encyclopedia of Human Computer Interaction (pp. 79-84). Hershey, PA: Idea
Group.
Akin, O., Baykan, C., & Rao, D. (1987). Structure of a directory space: A case study with
a UNIX operating system. International Journal of Man-Machine Studies, 26, 361- 382.
Ary, D. Jacobs, L. C., Razavieh, A. & Sorensen, C. K. (2009). Introduction to research in
education (8th ed.). Belmont, CA: Wadsworth Publishing.
Atwood, J. R., Hinds, P., Benoliel, J. Q. & Artinian, B. M. (1986). Heuristic heresy:
Application of reliability and validity criteria to products of grounded theory. Western
Journal of Nursing Research, 8(2), 135-154.
Barreau, D. K. (1995). Context as a factor in personal information management systems.
Journal of the American Society for Information Science, 46(5), 327-339.
Barreau, D. K. & Nardi, B. A. (1995). Finding and reminding: File organization from the
desktop. SIGCHI Bulletin, 27(3), 39-43.
156
Beazley, H., Boenisch, J. & Harden, D. (2002). Continuity Management: Preserving
Corporate Knowledge and Productivity When Employees Leave. Hoboken, N.J.: John
Wiley & Sons.
Becker, H. S. (1986). Writing for social scientists. Chicago: University of Chicago Press.
Bellotti, V. & Smith, I. (2000). Informing the design of an information management
system with iterative fieldwork. In Conference on Designing interactive systems (DIS
2000), New York City, New York, ACM Press.
Bergeron, B. (2003). Essentials of Knowledge Management. Hoboken, NJ: Wiley.
Bergman, O., Beyth-Marom, R. & Nachmias, R. (2008). The user-subjective approach to
personal information management systems design: Evidence and implementations.
Journal of the American Society for Information Science and Technology, 59(2), 235-246.
Bergman, O., Whittaker, S., Sanderson, M., Nachmias, R. & Ramamoorthy, A. (2010).
The effect of folder structure on personal file navigation. Journal of the American Society
for Information Science and Technology, 61(12), 2426-2441.
Bernstein, M., Van Kleek, M., Karger, D. Schraefel, M. (2008). Information scraps: How
and why information eludes our Personal Information Management tools. ACM
Transactions on Information Systems, 26(4), Article
24. Hhttp://people.csail.mit.edu/emax/papers/TOIS-a24-bernstein-vankleek.pdfH.
Bers, M. U. (2010). Beyond computer literacy: Supporting youth's positive development
through technology. New Directions for Youth Development, 128, 13-23.
Boardman, R. (2001a). Category Overlap between Hierarchies in User Workspace. In
Interact 2001, the Eighth IFIP TC 13 International Conference on Human- Computer
Interaction.
Boardman, R. (2001b). Multiple Hierarchies in user workspace. In Extended Abstracts of
CHI Conference on Human Factors in Computing Systems, ACM Press, Seattle.
157
Boardman, R., Spence, R., & Sasse, M.A. (2003). Too Many Hierarchies? The Daily
Struggle for Control of the Workspace, In Jacko, J., Stephanidis, C. (Eds.), Proceedings
of HCI International 2003 (Volume 1), 616-620.
Boardman, R. & Sasse, M.A. (2004). "Stuff goes into the computer and doesn't come out":
A Cross-tool Study of Personal Information Management. In Proceedings of CHI 2004,
ACM Conference on Human Factors in Computing Systems.
Boechler, P. M. (2006). Understanding Cognitive Processes in Educational Hypermedia.
In: Ghaoui, C. (Ed.). Encyclopedia of Human Computer Interaction (pp. 648-651).
Hershey, PA: Idea Group.
Bondarenko, O. & Janssen, R. (2005). Documents at Hand: Learning from Paper to
Improve Digital Technologies. Proceedings of CHI 2005, Portland, USA, April 2005,
121-130.
Borgman, C. L. (1984). Psychological research in human-computer interaction. In Annual
Review of Information Science and Technology, 19, 33-64.
Bowen, G. A. (2006). Grounded theory and sensitizing concepts. International Journal of
Qualitative Methods, 5(3), 12-23.
Broadbent, D. E. & Broadbent, M. H. P. (1978). The allocation of descriptor terms by
individuals in a simulated retrieval system. Ergonomics, 21(5), 343-354.
Bruce, H. (2005). Personal anticipated information need. Information Research,
10(3). Hhttp://informationr.net/ir/10-3/paper232.htmlH.
Burton, H. D. (1981). FAMULUS Revisited: Ten Years of Personal Information Systems.
Journal of the American Society for Information Science, 32(6), 440-443.
Bush, Vannevar. (1945). As We May Think. Atlantic Monthly, 176(1), 101-108.
158
Carroll, J. M. (1982). Creative names for personal files in an interactive computing
environment. International Journal of Man-Machine Studies, 16(4), 405-438.
Case, D. O. (1986). Collection and organization of written information by social
scientists and humanists: a review and exploratory study. Journal of Information Science,
12(3), 97-104.
Case, D. O. (1991). Conceptual organization and retrieval of text by historians: The role
of memory and metaphor. Journal of the American Society for Information Science, 42(9),
657-668.
Chapman, H. (1999). The file naming habits of personal computer users. Master's thesis,
University of North Carolina at Chapel Hill.
Charmaz, K. (2003). Grounded theory: Objectivist and constructivist methods. In N. K.
Denzin & Y.S. Lincoln (Eds.), Strategies for qualitative inquiry (2nd ed., pp. 249-291).
Thousand Oaks, CA: Sage.
Charmaz, K. (2006), Constructing Grounded Theory: A Practical Guide Through
Qualitative Analysis. London: Sage.
Chen C. (1974). How do scientists meet their information needs? Special Libraries, 65,
272-80.
Chiu, T. S. (2000). To promote the nurses’ competence in information application
through the life-long learning philosophy. VGH Nursing, 17(3), 248–259.
Cole, I. (1982). Human Aspects of Office Filing: Implications for the Electronic Office.
In Proceedings of the Human Factors Society 26th Annual Meeting, edited by Richard E.
Edwards, 59-63. Santa Monica: Human Factors Society.
159
Corbin, J. & Strauss, A. (2008). Basics of Qualitative Research. Thousand Oaks, CA:
Sage.
Czerwinski, M., Gage, D.W., Gemmell, J., Marshall, C.C., Perez-Quinonesis, M.A.,
Skeels, M.M., & Catarci, T. (2006, January). Digital Memories in an Era of Ubiquitous
Computing. Communications of the ACM, 49(1), 44-50.
Dalkir, K. (2005). Knowledge Management in Theory and Practice. Burlington, MA:
Elsevier Butterworth-Heinemann.
Dalkir, K. (2010). La continuité du savoir : préservation et transmission du savoir dans le
secteur public. Téléscope, 16(1), 146-167.
Davenport, T. H. & Pusak, L. (1998). Working Knowledge: How organizations manage
what they know. Boston, MA: Harvard Business School.
Davis, G. & Olson, M. (1985). Management Information Systems: Conceptual
Foundations, Structure and Development. New York: McGraw-Hill.
Dourish, P., Edwards, K., LaMarca, A. & Salisbury, M. (1999). Presto: An experimental
architecture for fluid interactive document spaces. ACM Transactions on Computer-
Human Interaction, 6(2), 133-161.
Drucker, P. F. (1973). Management: Tasks, Responsibilities and Practices, New York:
Harper & Row.
Ducheneaut, N. & Bellotti, V. (2001). E-mail as habitat: An exploration of embedded
personal information management. Interactions, 8(5), 30–38.
Dumais, S., Cutrell, E., Cadiz, J.J., Jancke, G., Sarin, R. & Robbins, D.C. (2003). Stuff
I've Seen: A system for personal information retrieval and re-use. In 26th Annual
160
International ACM SIGIR Conference on Research and Development in Information
Retrieval (SIGIR 2003).
Durst, S. (1994, March 27). It’s Time to Do Something About This Sexist Term
“Secretary”. The New York Times. Hhttp://www.nytimes.com/1994/03/27/opinion/l-it-s-
time-to-do-something-about-this-sexist-term-secretary-589357.html
Edmunds, A. & Morris, A. (2000). The problem of information overload in business
organisations: a review of the literature. International Journal of Information
Management, 20(1), 17-28.
Engelbart, D. C. (1961). Special considerations of the individual as a user, generator, and
retriever of information. American Documentation, 12, 121-125.
Eppler, M.J. & Mengis, J., (2003). A Framework for Information Overload Research in
Organizations: Insights from Organization Science, Accounting, Marketing, MIS, and
Related Disciplines. Working Paper #1/2003 ICA, University of Lugano, Lugano.
Erlandson, D. A., Harris, E. L., Skipper, B. L. & Allen, S. D. (1993). Doing naturalistic
inquiry: A guide to methods. Newbury Park, CA: Sage.
Ferraioli, L. (2005). An Exploratory Study of Metadata Creation in a Health Care Agency.
Cataloging & Classification Quarterly, 40(3-4), 75-102.
Fertig, S., Freeman, E. & Gelernter, D. (1996a). “Finding and reminding” reconsidered.
SIGCHI Bulletin, 28(1), 66-69.
Fertig, S., Freeman, E., & Gelernter, D. (1996b) Lifestreams: an alternative to the
desktop metaphor. Proceedings of the SIGCHI conference on Human factors in
computing systems, 410-411.
Field, A. (2003). When employees leave the company, how can you make sure that their
expertise doesn’t? Harvard management Communication Letter, 6(4), 3-5.
161
Figuerola, C., Rodríguez, A. & Berrocal, J. (2001). Automatic vs. manual categorization
of documents in Spanish. Journal of Documentation, 57, 763-773.
Floridi, L. (2005). Semantic Conceptions of Information. In: Zalta, E. (ed.), The Stanford
Encyclopedia of Philosophy (Winter 2005 Edition).
Hhttp://plato.stanford.edu/entries/information-semantic/H.
Gemmell, J., Bell, G., Lueder, R., Drucker, S., & Wong, C. (2002). Mylifebits: Fulfilling
the memex vision. Proceedings of the 2002 ACM workshops on Multimedia, 235–238.
Ghaoui, C. (Ed.). (2006). Encyclopedia of Human Computer Interaction. Hershey, PA:
Idea Group.
Giske, T. & Artinian, B. (2007). A personal experience of working with classical
grounded theory: From beginner to experienced grounded theorist. International Journal
of Qualitative Methods, 6(4), 67-80.
Glaser, B. G. (1978). Theoretical sensitivity: Advances in the methodology of grounded
theory. Mill Valley, CA: Sociology Press.
Glaser, B. G. (1998). Doing grounded theory: Issues and discussions. Mill Valley, CA:
Sociology Press.
Glaser, B. G. & Strauss, A. L. (1967). The discovery of grounded theory: Strategies for
qualitative research. Hawthorne, NY: Aldine de Gruyter.
Gonçalves, D. & Jorge, J. A. (2003). An Empirical Study of Personal Document Spaces.
Proceedings DSV-IS'03, 6-9 June 2003, Funchal, Portugal, 47-60. Hhttp://vimmi.inesc-
id.pt/~djvg/research/publications/documents/empirical_study_pds.pdf
162
Gordon, M. (1997). It's 10 a.m. do you know where your documents are? The nature and
scope of information retrieval problems in business. Information Processing &
Management 33(1): 107-122.
Gottlieb, L. & Dilevko, J. (2001). User preferences in the classification of electronic
bookmarks: Implications for a shared system. Journal of the American Society for
Information Science and Technology, 52(7), 517-535.
Gottlieb, L. & Dilevko, J. (2003). Investigating How Individuals Conceptually and
Physically Structure File Folders for Electronic Bookmarks: The Example of the
Financial Services Industry. Journal of the American Society for Information Science and
Technology, 54(2), 124-139.
Gourlay, S. N. (2002). Tacit knowledge, tacit knowing or behaving? 3rd European
Organizational Knowledge, Learning, and Capabilities conference, Athens, Greece, 5-6
April. Hhttp://myweb.tiscali.co.uk/sngourlay/PDFs/Gourlay%202002%20tacit%20knowled
ge.pdf
Gwizdka, J. (2002). TaskView: Design and evaluation of a task-based email interface.
Proceedings of the IBM Centers for Advanced Studies Conference - CASCON 2002, 136-
145.
Gwizdka, J. (2004). Email task management styles: The cleaners and the keepers. CHI
'04 Extended Abstracts on Human Factors in Computing Systems. Vienna, Austria. April
24-29, 2004, 1235-1238.
Gwizdka, J. & Chignell, M. H. (2004). Individual differences and task-based user
interface evaluation: A case study of pending tasks in email. Interacting with Computers,
16(4), 769-797.
163
Gwizdka, J. & Chignell, M. H. (2007). Individual Differences. In W. Jones & J. Teevan
(Eds.), Personal Information Management (pp. 206-220). Seattle: University of
Washington Press.
Hale, E. D., Treharne, G. J. & Kitas, G. D. (2007). Qualitative methodologies I: Asking
research questions with reflexive insight. Musculoskeletal Care, 5(3), 139–147.
Hammond, T., Hannay, T, Lund, B. & Scott, J. (2005). Social Bookmarking Tools (I): A
General Review. D-Lib Magazine, 11(4).
Hhttp://www.dlib.org/dlib/april05/hammond/04hammond.html H.
Heath, H. & Cowley, S. (2004). Developing a grounded theory approach: a comparison
of Glaser and Strauss. International Journal of Nursing Studies, 41, 141-150.
Henderson, S. (2004). How do people organize their desktops? CHI '04 extended
abstracts on Human factors in computing systems, 1047-1048.
Henderson, S. (2005). Genre, Task, Topic and Time: Facets of Personal Digital
Document Management. Proceedings of the 6th ACM SIGCHI New Zealand chapter's
international conference on Computer-human interaction, 75-82.
Henderson, S. (2009). How Do People Manage Their Documents? : An Empirical
Investigation Into Personal Document Management Practices Among Knowledge
Workers. Ph.D. Thesis, University of Auckland.
Hhttp://researchspace.auckland.ac.nz/handle/2292/5230
Hertzum, M. (1999). Six roles of documents in professionals’ work. In S. Bødker, M.
Kyng, and K. Schmidt (eds.), ECSCW’99: Proceedings of the Sixth European conference
on Computer Supported Cooperative Work (Copenhagen, Denmark, September 12-16,
1999), 41-60. Dordrecht: Kluwer.
164
Hey, J. (2004). The Data, Information, Knowledge, Wisdom Chain: The Metaphorical
Link. Hhttp://ioc.unesco.org/Oceanteacher/OceanTeacher2/02_InfTchSciCmm/DIKWchain
.pdfH.
Hjorland, B. (2003). Fundamentals of knowledge organization. Knowledge Organization,
30(2), 87-111.
Hoffmann, E. (1980). Defining Information: An Analysis of the Information Content of
Documents. Information Processing & Management, 16, 291-304.
Indratmo, I. & Vassileva, V. (2006). No More Isolated Files: Managing Files As Social
Artifacts. CHI Extended Abstracts 2006, 899-904.
Jacob, E. K. (2001). The everyday world of work: Two approaches to the investigation of
classification in context. Journal of Documentation, 57(1), 76-99.
Jafari, M., Akhavan, P. Bouroni, Atieh, B. & Amiri, R. H. (2009). A Framework For The
Selection Of Knowledge Mapping Techniques. Journal of Knowledge Management
Practice, 10(1). Hhttp://www.tlainc.com/articl180.htmH.
Jahoda, G. E., Hutchens, R.D. & Galford, R.R. (1966). Characteristics and Use of
Personal Indexes Maintained by Scientists and Engineers in One University. American
Documentation, 17, 71-75.
Jones, W. (2004). Finders, keepers? The present and future perfect in support of personal
information management. First Monday,
9(3). Hhttp://firstmonday.org/issues/issue9_3/jones/index.html
Jones, W. (2007). How People Keep and Organize Personal Information. In Jones, W. &
Teevan, J. Personal Information Management. 35-56.
Jones, W. (2008). Keeping Found Things Found: The Study and Practice of Personal
Information Management. Burlington, MA: Morgan Kaufmann.
165
Jones, W., Dumais, S. & Bruce, H. (2002). Once found, what then? : A study of
"keeping" behaviors in the personal use of web information. In 65th Annual Meeting of
the American Society for Information Science and Technology (ASIST 2002),
Philadelphia, PA, American Society for Information Science & Technology.
Jones, W., Phuwanartnurak A. J., Gill R. & Bruce, H. (2005). Don't take my folders
away! : Organizing personal information to get things done. CHI '05 extended abstracts
on Human factors in computing systems, April 02-07, 2005, Portland, OR, USA, p. 58-64.
Jones, W. & Teevan, J. (2007). Personal Information Management. Seattle: University
of Washington Press.
Kalid, K. S. B. & Mahmood, A. K. B. (2010). Using Stories To Share Knowledge: A
Malaysian Organization Case Study. Journal of Knowledge Management Practice,
11(1). Hhttp://www.tlainc.com/articl222.htmH.
Karger, D. R., Bakshi, K., Huynh, D., Quan, D. & Sinha, V. (2003). Haystack: A
customizable general-purpose information management tool for end users of
semistructured data. Proceedings of the 2003 CIDR Conference. Hhttp://www-
db.cs.wisc.edu/cidr/cidr2005/papers/P02.pdfH.
Karger, D. & Jones, W. (2006). Data unification in personal information management.
Communications of the ACM, 49(1), 77-82.
Kelly, D. & Teevan, J. (2007). Understanding what works: Evaluating PIM tools. In W.
Jones & J. Teevan (Eds.), Personal Information Management (pp. 190-204). Seattle:
University of Washington Press.
Khoo, C., Luyt, B., Ee, C., Osman, J., Lim, H.H. & Yong, S. (2007). How Users
Organize Electronic Files on Their Workstations in the Office Environment: A
Preliminary Study of Personal Information Organization Behaviour. Information
Research, 12(2). Hhttp://InformationR.net/ir/12-2/paper293.htmlH.
166
Kidd, A. (1994). The marks are on the knowledge worker. Proceedings of CHI ’94, 186-
191.
Kwasnik, B. H. (1991). The importance of factors that are not document attributes in the
organization of personal documents. Journal of Documentation, 47(4), 389-398.
Lansdale, M. (1988). The psychology of personal information management. Applied
Ergonomics, 19(1), 55-66.
Lansdale, M. W. (1991). Remembering about documents : memory for appearance,
format, and location. Ergonomics, 34(8), 1161-1178.
Lave, J. & Wenger, E. (1991). Situated Learning: Legitimate Peripheral Participation.
Cambridge: Cambridge University Press.
Levien, R. E. (1989). The Civilizing Currency: Documents and Their Revolutionary
Technologies. Stanford, CT: Xerox Corporation.
Lin J., Lin K., Jiang W & Lee T. (2007). An exploration of nursing informatics
competency and satisfaction related to network education. Journal of Nursing Research,
15(1), 54-65.
Lincoln, Y. S. & Guba, E. G. (1985). Naturalistic inquiry. Beverly Hills: Sage
Publications.
Lomborg, K. & Kirkevold, M. (2003). Truth and validity in grounded theory – a
reconsidered realist interpretation of the criteria: fit, work, relevance and modifiability.
Nursing Philosophy. 4, 189-200.
Mackenzie, M. L. (2000). The Personal Organization of Electronic Mail Messages in a
Business Environment: An Exploratory Study. Library & Information Science Research,
22(4), 405-426.
167
Mackenzie, M. L. (2002). Storage and retrieval of e-mail in a business environment: An
exploratory study. Library & information science research, 24(4), 357-372.
Malone, T. W. (1983). How do people organize their desks: implications for the design of
office information-systems. ACM Transactions on Office Information Systems, 1(1), 99-
112.
Mander, R., Salomon, G. & Wong, Y. Y. (1992). A "pile" metaphor for supporting casual
organization of information. CHI 92: ACM SIGCHI Conference on Human Factors in
Computing, Monterey, CA, 627-634.
Mansourian, Y. (2006). Adoption of grounded theory in LIS research. New Library
World, 107 (1228/1229), 386-402.
Mariampolski, H. (2001). Qualitative Market Research. Thousand Oaks, CA: Sage
Publishing.
Mas, S. (2007). Schémas de classification et repérage des documents administratifs
électroniques dans un contexte de gestion décentralisée des ressources informationnelles.
Ph.D. Thesis, Université de Montréal.
Maxwell, J. A. (1996). Qualitative research design: An interactive approach. Thousand
Oaks, CA: Sage.
McIlwaine, I.C. (2003). Trends in knowledge organization research. Knowledge
Organization, 30(2), 75-86.
Mehra, B. (2002). Bias in Qualitative Research: Voices from an Online Classroom. The
Qualitative Report, 7(1). Hhttp://www.nova.edu/ssss/QR/QR7-1/mehra.html H.
Meier, J. & Sprague, R. (1996). Towards a Better Understanding of Electronic Document
Management. 29th Annual Hawaii International Conference on System Sciences (HICSS-
29), January 3-6, 1996, Maui, Hawaii. IEEE Computer Society, 5, 53-61.
168
Modjeska, D. & Chignell, M. (2003). Individual Differences in Exploration Using
Desktop VR. Journal of the American Society for Information Science and Technology,
54(3), 216-228.
Murphy, J. E. (1980). Filing of Personal Reference Collections. American Journal of
Hospital Pharmacy, 37(5), 618.
Najarian, S. E. (1981). Organizational factors in human memory: implications for library
organization and access systems. Library Quarterly, 51(3), 269-291.
Nardi & Barreau (1997). “Finding and reminding” revisited: Appropriate metaphors for
file organization at the desktop. SIGCHI Bulletin, 29(1), 76-78.
Naumer, C. M. & Fisher, K. E. (2007). Naturalistic approaches for understanding PIM. In
W. Jones & J. Teevan (Eds.), Personal Information Management (pp. 76-88). Seattle:
University of Washington Press.
Neumann, L. J. (1999). Paper, piles, and computer files: folklore of information work
environments. Library Trends, 47(3), 439-469.
Neustaedter, C., Brush, A.J. & Smith, M.A. (2005). Beyond “From” and “Received”:
Exploring the dynamics of email triage. CHI Extended Abstracts 2005, 1977-1980.
Nonaka, I. (1994). A dynamic theory of organizational knowledge creation. Organization
Science, 5 (1), 14-37.
Norman, D. A. (1988). The psychology of everyday things. New York: Basic Books.
Paré, F.-X. (2006). L'importance relative des documents : impacts sur l'organisation
personnelle de l’information en milieu de travail. Unpublished paper presented at the
74ème Congrès de l’ACFAS, McGill University.
Paré, F.-X. (2007). The many facets of document importance: A case study of office
workers. CAIS 2007, McGill University.
169
Hhttp://www.cais-acsi.ca/proceedings/2007/pare_2007.pdfH.
Parker, E.B. & Paisley, W.J. (1966). Research for psychologists at the interface of the
scientist and his information system. American Psychologist, 21(11), 1061–1071.
Pédauque, R. T. (2003). Document: Form, Sign and Medium, As Reformulated for
Electronic Documents. Working paper, version 3, July 8. STIC-CNRS.
Hhttp://archivesic.ccsd.cnrs.fr/documents/archives0/00/00/05/94/sic_00000594_01/sic_000
00594.pdfH.
Pettigrew, K. E., Fidel, R. & Bruce, H. (2001). Conceptual frameworks in information
behavior. Annual Review of Information Science and Technology, 35, 43-78
Pratt, W., Unruh, K., Civan, A., & Skeels, M. (2006). Managing Health Information in
Your Life. Communications of ACM, 49(1), 51-55.
Pulakos, E. D., Arad, S., Donovan, M. & Plamondon, K. E. (1999). Adaptability in the
Workplace: Development of a Taxonomy of Adaptive Performance. 23rd Annual
IPMAAC Conference on Professional Personnel Assessment, St. Petersburg Beach,
Florida, June 7-11, 1999. Hhttp://www.ipacweb.org/conf/99/pulakos.pdf
Punch, K. F. (2000). Developing effective research proposals. Thousand Oaks, CA: Sage.
Ravasio P., Schär, S. G. & Kruger, H. (2004). In Pursuit of Desktop Evolution: User
Problems and Practices with Modern Desktop Systems. ACM Transactions on Computer-
Human Interaction, 11(2), 156-180.
Ringel, M., Cutrell, E., Dumais, S. & Horvitz, E. (2003). Milestones in time: The value of
landmarks in retrieving information from personal stores. Proceedings of INTERACT
2003, 184-191.
Rosch, E. (1978). Principles of categorization. Hillsdale, NJ: Erlbaum.
170
Rousseau, J.-Y. & Couture, C. (1994). Les fondements de la discipline archivistique. Ste-
Foy, Presses de l'Université du Québec.
Schamber, L. (1996). What is a Document? Rethinking the Concept in Uneasy Times.
Journal of the American Society for Information Science, 47(9), 669-671.
Seldén, L. (2005). On grounded theory – with some malice. Journal of Documentation,
61(1), 114-129.
Soper, M. E. (1976). Characteristics and use of personal collections. Library Quarterly,
46(4), 397-415.
Sprague, R. (1995). Electronic document management: Challenges and opportunities for
information systems managers. MIS Quarterly, 19(1), 29-50.
Stern, P. N. (1995). Grounded Theory Methodology: Its Uses and Processes. In
Grounded Theory 1984-1994 (Glaser, B. G., ed.), Mill Valley, CA: Sociology Press.
Strauss, A. L. & Corbin, J. (1990). Basics of Qualitative Research: Techniques and
Procedures for Developing Grounded Theory. Thousand Oaks, CA: Sage.
Svenonius, E. (2000). The intellectual foundation in information organization.
Cambridge, Mass.: MIT Press.
Thomson, S. B. (2004). Qualitative Research: Grounded Theory : Sample Size and
Validity. Hhttp://www.buseco.monash.edu.au/research/studentdocs/mgt.pdfH.
Trochim, W. M. (2006). Qualitative validity. The Research Methods Knowledge Base,
2nd Edition. Hhttp://www.socialresearchmethods.net/kb/qualval.phpH.
Tyrväinen, P. & Päivärinta, T. (1999). On Rethinking Organizational Document Genres
for Electronic Document Management. 32nd Annual Hawaii International Conference on
System Sciences (HICSS-32), 5-8 January, 1999, Maui, Hawaii, Track 2: Digital
Documents. IEEE Computer Society.
171
Hhttp://computer.org/proceedings/hicss/0001/00012/0001toc.htmH.
Tsai, M.-J. (2002). Do male students often perform better than female students when
learning computers?: A study of Taiwanese eight graders' computer education through
strategic and cooperative learning. Journal of Educational Computing Research, 26(1),
67-85.
Urquhart, C. & Fernandez, W. (2006). Grounded theory method: The researcher as blank
slate and other myths. ICIS 2006 Proceedings, Paper
31. Hhttp://aisel.aisnet.org/icis2006/31H.
U.S. Bureau of Labor Statistics. (2009). Occupational Outlook Handbook, 2010-11
Edition: Secretaries and Administrative Assistants. Hhttp://www.bls.gov/oco/ocos151.htmH.
Weed, M. (2009). Research quality considerations for grounded theory research in sport
& exercise psychology. Psychology of Sport and Exercise, 10, 502-510.
Whittaker, S. & Hirschberg, J. (2001). The character, value and management of paper
archives. ACM Transactions on Computer Human Interaction, 8, 150-170.
Whittaker, S. & Sidner, C. (1996). Email overload: Exploring personal information
management of email. In Conference on Human Factors in Computing Systems (CHI
1996), Vancouver, B.C., ACM SIGCHI.
Wilson, T.D. (1999). Models in information behaviour research. Journal of
Documentation, 55, 249-270.
Yiu, K.S., Baecker, R., Silver, N. & Long, B. (1997). A time-based interface for
electronic mail and task management. Proceedings of HCI International 1997, 2, 19-22.
172
9B119BAppendix 1: Example of an e-mail invitation to participate in the study
Bonjour,
Mon nom est François-Xavier Paré, et j'étudie présentement au doctorat en Sciences de
l'information à l'Université McGill. Mes recherches portent sur la façon dont les gens
organisent l'information sur leur ordinateur au bureau. Je suis également employé de la
Direction des bibliothèques de l’Université de Montréal.
Je vous écris aujourd'hui parce que j'aimerais beaucoup vous interviewer dans le cadre de
ma recherche. Les questions porteraient sur la façon dont vous classez vos documents,
pourquoi vous organisez les choses de telle ou telle façon, etc. Le beau côté, c'est qu'il n'y
a pas de bonnes ou de mauvaises réponses…
Si vous acceptez, vous auriez d’abord à remplir un petit questionnaire en ligne (10-15
minutes). Il y aurait ensuite une entrevue qui pourrait se dérouler dans votre bureau et
durerait probablement entre 45 minutes et une heure. Je vous laisserais évidemment
choisir le moment le plus approprié ; je suis assez flexible de ce côté-là. L'entrevue serait
enregistrée, mais le tout demeure totalement confidentiel (i.e. votre nom n'apparaîtra
nulle part), alors pas d'inquiétude à avoir de ce côté. Il serait quand même préférable
d'obtenir l'aval de votre patron(ne) avant de procéder.
Évidemment, rien ne vous oblige à participer, mais ça me rendrait vraiment un très grand
service... Si vous n'êtes pas certaine ou si vous avez des questions, n'hésitez pas à
communiquer avec moi.
Merci !
François-Xavier Paré
Candidat au doctorat, School of Information, Université McGill
Bibliothécaire, Direction des bibliothèques, Université de Montréal
173
10B120BAppendix 2: Interview Guide
A) Signing of the consent form (2 copies)
B) Start recording
C) Interview questions
1) Nature of the job:
- Main tasks and responsibilities, mission in the organization, main projects
- Place in the organization (superiors, colleagues, interactions with whom, place
in the organizational chart)
2) Documents:
- Documents typically produced
- Documents received: from whom?
- Documents sent: to whom?
- Documents used
3) Document organization:
- Organizing principles, objectives
- Documents inherited from the predecessor, transition.
- Local rules for organizing or conserving documents.
- Additions and modifications to the structure (rhythm).
- Paper documents, e-mail, other files (same structure?)
174
- Printing habits.
- Shared folders (issues?)
- Personal opinion on organizing documents, finding documents, etc.
- Backup procedure
4) Importance / Value of documents:
- Important documents and why.
- Less important documents and why.
D) Screen capture(s) of folder structure in Windows Explorer (Mes Documents, Windows
Desktop, and/or other relevant folders) – sent via e-mail or saved on USB key.
175
11B121BAppendix 3: Pre-interview survey
A. Accueil
Gestion personnelle de l’information en milieu de travail
Enquête sur les pratiques d'employés de l'Université X
Ce questionnaire fait partie d'une recherche doctorale sur l’organisation des documents
électroniques en milieu de travail. Le but du questionnaire est de recueillir des
informations préalables afin de mieux préparer l’entrevue qui sera menée avec vous sous
peu. Le questionnaire devrait prendre une vingtaine de minutes à compléter.
Soyez assuré(e) que vos réponses demeureront strictement confidentielles.
Merci !
François-Xavier Paré
Candidat au doctorat, School of Information, Université McGill
Téléphone : (XXX) XXX-XXXX, poste XXXX
Courriel : XXXX
176
B. Informations générales
1. Quel est votre nom ?
2. Dans quel département/faculté/école/service travaillez-vous (ex. : Bureau du
registraire) ?
3. Depuis combien d’années travaillez-vous à l’Université X?
Moins de 1 an
1 à 5 ans
6 à 10 ans
11 à 15 ans
16 ans et +
4. Depuis combien d’années travaillez-vous dans ce département/faculté/école/service ?
Moins de 1 an
1 à 5 ans
6 à 10 ans
11 à 15 ans
16 ans et +
5. Quel est le titre du poste actuellement occupé ?
177
6. Depuis combien d’années occupez-vous ce poste ?
Moins de 1 an
1 à 5 ans
6 à 10 ans
11 à 15 ans
16 ans et +
7. Quels sont, parmi les énoncés suivants, ceux qui correspondent à vos tâches ? (merci
de cocher tous les énoncés qui s'appliquent)
Je participe à différentes opérations reliées à la gestion des études
Je participe à différentes opérations reliées à la gestion des ressources humaines
Je participe à différentes opérations reliées à la gestion des ressources financières
Je participe à différentes opérations reliées à la gestion des ressources matérielles
J’effectue le suivi des décisions prises par mon ou ma supérieure et prend les
mesures appropriées afin d'en assurer la réalisation.
Je prends les dispositions relatives à l'agenda de mon ou ma supérieure. À cette fin,
je prends et fixe les rendez-vous, les lui rappelle et organise des rencontres selon les
priorités.
Je tiens à jour la comptabilité de l’unité. J’enregistre les dépenses effectuées et fait,
sur demande, un rapport des états de compte. Je compare les écritures comptables avec
les pièces justificatives sur réception périodique des relevés budgétaires.
Je tiens à jour les dossiers du personnel et prépare les documents requis lors des
affichages, des engagements, des promotions, des nominations, des affectations, des
congés sabbatiques, des absences, etc.
J’accomplis les tâches de la technicienne ou du technicien à la gestion des dossiers
178
étudiants lorsque dans le secteur de travail il n’y a pas de poste de cette nature ou que la
charge de travail de la technicienne ou du technicien à la gestion des dossiers étudiants le
requiert.
Je reçois et filtre les appels téléphoniques et donne des renseignements. J’accueille
les visiteuses et les visiteurs, m’informe de l’objet de l’entretien et prépare les documents
nécessaires. J’oriente ceux dont le problème peut être réglé par une autre autorité et fixe
des rendez-vous si, à mon avis, il est nécessaire de tenir un entretien avec mon ou ma
supérieure.
Je prends connaissance du courrier et le trie en fonction des priorités. Je prépare les
dossiers en faisant les recherches et les consultations préalables. Je réponds aux
demandes relevant de ma compétence. Je rédige des projets de réponse complexes. Je
note les suites à donner et effectue le suivi des dossiers.
J’organise des réunions en faisant les convocations, les réservations des salles et la
préparation du matériel et des documents nécessaires. J’assiste aux réunions, prends note
des délibérations, rédige les procès-verbaux et donne suite aux décisions qui relèvent de
ma compétence.
J’assume la responsabilité du système de classement et applique la procédure
d’archivage. J’effectue la mise à jour régulière des dossiers et des fichiers informatiques
de gestion.
Je prends des notes, dactylographie divers documents tels que lettres, rapports,
mémos, etc. J’effectue la mise en page des documents et en vérifie l’orthographe et la
grammaire. Je signe certains documents.
J’utilise les logiciels en usage. Je tiens à jour les applications informatisées courantes
et participe à leur développement afin d’améliorer l’efficacité des opérations.
Je participe à la planification et à l’organisation matérielle d’activités spéciales telles
que colloques, congrès, voyages.
J’accomplis temporairement les tâches d’un poste connexe ou inférieur lorsque
requis.
179
Autres (merci de spécifier) :
C. Compétences en informatique
1. Depuis combien d’années utilisez-vous un ordinateur ?
Moins de 1 an
1 - 5 ans
6 - 10 ans
Plus de 10 ans
2. Quel système d’exploitation utilisez-vous principalement dans votre milieu de travail ?
Windows
Mac OS
Unix
Je ne sais pas
Autre système d'exploitation (merci de préciser) :
3. Parmi les descriptions des niveaux de compétence en informatique qui suivent, cochez
l’énoncé qui correspond le mieux à votre niveau.
Novice : J’essaie de me servir de l’informatique mais j’ai souvent besoin d’aide.
Intermédiaire : Je peux exécuter des tâches simples avec un nombre limité de
logiciels.
180
Expert : Je peux utiliser d’une façon compétente une grande variété de technologies
informatiques.
D. Formation en gestion des documents administratifs
La gestion des documents administratifs désigne l'ensemble des activités et méthodes
relatives à la mise en place de systèmes d'administration des documents permettant
d'organiser, de décrire, de conserver et de retrouver, sûrement et à tout moment, tout
document produit ou reçu dans le cadre des activités d'un organisme.
1. Avez-vous déjà reçu la formation en gestion des documents administratifs donnée par
la Division des archives de l'Université X?
Oui
Non
Si oui, indiquer en quelle année. :
2. Avez-vous déjà reçu une formation en gestion des documents administratifs donnée
dans le cadre d'un programme d'études collégiales (i.e. niveau Cégep) ?
Oui
Non
Si oui, veuillez spécifier l'année ainsi que le nom du programme et l'établissement.
3. Avez-vous déjà reçu une formation en gestion des documents administratifs donnée
dans le cadre d'un programme d'études universitaires ?
Oui
181
Non
Si oui, veuillez spécifier l'année ainsi que le nom du programme et l'établissement.
4. Est-ce que l'une ou l'autre des formations suivies comprenait la classification des
documents ?
Oui
Non
Je ne m'en souviens pas
5. Est-ce que l'une ou l'autre des formations suivies comprenait la classification des
"documents électroniques" ?
Oui
Non
Je ne m'en souviens pas
E. Plan de classification institutionnel
Certain(e)s employé(e)s de l'Université X utilisent le plan de classification institutionnel
proposé par la Division des archives de l'Université X pour concevoir cette structure de
répertoires. D'autres employé(e)s conçoivent une structure de répertoires plus
personnalisée. Les questions qui suivent visent à mieux connaître votre propre structure
de répertoires.
1. Utilisez-vous le plan de classification institutionnel conçu par la Division des archives
182
de l’Université X pour organiser et nommer votre structure de répertoires se trouvant sur
votre espace personnel de travail ?
Oui, j'utilise intégralement le plan de classification institutionnel pour organiser mes
documents électroniques. Allez à la section F.
Oui, j'utilise en partie le plan de classification institutionnel pour organiser mes
documents électroniques. Allez à la section F.
Non, je n'utilise pas du tout le plan de classification institutionnel pour organiser mes
documents électroniques. Allez à la section G.
Je ne sais pas. Allez à la section H.
F. Plan de classification institutionnel (suite)
1. Pourquoi utilisez-vous le plan de classification institutionnel ? (cocher tous les énoncés
qui s'appliquent)
Parce que le plan de classification était déjà installé sur l’ordinateur qui m’a été
alloué
Parce que l’utilisation du plan de classification a été rendue obligatoire par mon ou
ma supérieure
Pour faciliter l’organisation de mes documents électroniques
Pour faciliter le repérage de mes documents électroniques
Comme support à la gestion de mes dossiers
Autre(s) raison(s) :
183
G. Plan de classification institutionnel (suite)
1. Qu’est-ce qui vous retient d’utiliser le plan de classification institutionnel ? (cocher
tous les énoncés qui s'appliquent)
Je n’ai pas eu connaissance de cet outil
Il est trop complexe/difficile à utiliser
Il n’est pas assez précis
Il est trop détaillé
J’utilise déjà mon propre plan de classification
Autre(s) raison(s) :
2. Qu'est-ce qui vous motiverait à utiliser le plan de classification institutionnel ?
H. Auto-évaluation de l’organisation des documents électroniques
1. Dans l’ensemble, comment qualifieriez-vous l’organisation de vos documents
électroniques ?
Documents très désorganisés
Documents désorganisés
184
Documents moyennement organisés
Documents bien organisés
Documents très bien organisés
2. Dans l’ensemble, comment évaluez-vous le niveau de difficulté que comporte
l’organisation de vos documents électroniques ?
Tâche très difficile
Tâche difficile
Tâche moyennement facile
Tâche facile
Tâche très facile
3. À quelle fréquence vous arrive-t-il de remarquer qu’un document électronique a été
placé dans le mauvais répertoire ?
Jamais
1 à 2 fois par année
Au moins 1 fois par session
1 à 2 fois par mois
1 à 2 fois par semaine
Au moins une fois par jour
Je ne sais pas
4. À quelle fréquence vous arrive-t-il de NE PAS être capable de retrouver un document
électronique ?
Jamais
185
1 à 2 fois par année
Au moins 1 fois par session
1 à 2 fois par mois
1 à 2 fois par semaine
Au moins une fois par jour
Je ne sais pas
5. Dans l’ensemble, comment qualifieriez-vous le repérage de vos documents
électroniques sur votre ordinateur ?
Repérage très difficile
Repérage difficile
Repérage moyennement facile
Repérage facile
Repérage très facile
I. Importance des documents
Il y a probablement dans votre ordinateur des documents plus importants que d’autres, et
ce pour diverses raisons. En prévision de l'entrevue, veuillez identifier 3 documents que
vous jugez PLUS IMPORTANTS, parmi ceux qui se trouvent sur votre poste, ainsi que 3
documents que vous jugez MOINS IMPORTANTS.
Indiquez pour chacun des documents son nom (ex.: Résumé.doc), une brève description
de son contenu (ex.: Résumé des trois derniers exercices financiers), ainsi que sa
localisation sur votre ordinateur (ex.: Mes Documents/Budget/Résumé.doc).
1. Premier document jugé IMPORTANT :
186
Nom du document :
Description du document :
Localisation du document :
2. Deuxième document jugé IMPORTANT :
Nom du document :
Description du document :
Localisation du document :
3. Troisième document jugé IMPORTANT :
Nom du document :
Description du document :
Localisation du document :
4. Premier document jugé PEU IMPORTANT :
Nom du document :
Description du document :
Localisation du document :
5. Deuxième document jugé PEU IMPORTANT :
Nom du document :
Description du document :
Localisation du document :
187
6. Troisième document jugé PEU IMPORTANT :
Nom du document :
Description du document :
Localisation du document :
J. Fin
Vous avez terminé de remplir le questionnaire !
Un grand merci !
François-Xavier Paré
188
12B122BAppendix 4: Participant profiles
189
42B152BParticipant #1: Anne
Age: 30-39 Years at the university: 6-10 Position: Office Work Technician (OWT). Years in this position: 1-5 Use of the institutional scheme: No. Training in document management: Secretary college. Self-assessment of computer literacy: Expert. Self-assessment of current document organization: Very well organized. Level of difficulty in organizing documents: Very Easy. Main tasks:
- Participates in operations pertaining to the management of financial resources. - Participates in operations pertaining to the management of material resources. - Follows up on the decisions taken by his/her supervisor and makes sure things get
done. - Manages his/her superior's agenda (fixes appointments according to priorities). - Manages the unit's accounts, registers expenses, produces reports, etc. - Receives and filters phone calls, gives out information. Welcomes visitors and
prepares necessary documents. - Sorts incoming mail according to priorities. Prepares responses and takes note of
any necessary follow-up. - Organizes meetings by sending out invitations, booking meeting rooms and
preparing the required documents. Takes part in meetings, takes notes, and ensures any necessary follow-up.
- Manages the local filing system and applies the required archival procedures. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents. - Participates in the planning of special events such as conferences or travels. - Accomplishes the tasks of a similar or lower position when needed. - Writes and/or translates documents when needed (English/French).
190
43B153BParticipant #2: Béatrice
Age: 40-49 Years at the university: 6-10 Position: Secretarial Coordinator (SC). Years in this position: 6-10 Use of the institutional scheme: No. Training in document management: In Cégep; also received training from the
University’s Archives Division. Self-assessment of computer literacy: Intermediate. Self-assessment of current document organization: Well organized. Level of difficulty in organizing documents: Moderately Easy. Main tasks:
- Participates in operations pertaining to the management of human resources. - Follows up on the decisions taken by his/her supervisor and makes sure things get
done. - Manages his/her superior's agenda (fixes appointments according to priorities). - Receives and filters phone calls, gives out information. Welcomes visitors and
prepares necessary documents. - Sorts incoming mail according to priorities. Prepares responses and takes note of
any necessary follow-up. - Organizes meetings by sending out invitations, booking meeting rooms and
preparing the required documents. Takes part in meetings, takes notes, and ensures any necessary follow-up.
- Takes notes and types various documents such as letters, reports, memos, etc. Prepares page layout and verifies grammar and spelling. Signs certain documents.
- Accomplishes the tasks of a similar or lower-ranked position when needed.
191
44B154BParticipant #3: Carole
Age: 40-49 Years at the university: 6-10 Position: Secretarial Agent (SA). Years in this position: 1-5 Use of the institutional scheme: No. Training in document management: Received training from the University’s Archives
Division. Self-assessment of computer literacy: Intermediate. Self-assessment of current document organization: Well organized. Level of difficulty in organizing documents: Moderately Easy. Main tasks:
- Follows up on the decisions taken by his/her supervisor and makes sure things get done.
- Manages the local filing system and applies the required archival procedures. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents.
192
45B155BParticipant #4: Diane
Age: 40-49 Years at the university: 16+ Position: Office Work Technician (OWT). Years in this position: 0-1 Use of the institutional scheme: No. Training in document management: Received training from the University’s Archives
Division. Self-assessment of computer literacy: Expert. Self-assessment of current document organization: Moderately satisfied. Level of difficulty in organizing documents: Easy. Main tasks:
- Follows up on the decisions taken by his/her supervisor and makes sure things get done.
- Sorts incoming mail according to priorities. Prepares responses and takes note of any necessary follow-up.
- Manages the local filing system and applies the required archival procedures. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents.
193
46B156BParticipant #5: Édith
Age: 30-39 Years at the university: 1-5 Position: Office Work Technician (OWT). Years in this position: 1-5 Use of the institutional scheme: No. Training in document management: In Cégep; also received training from the
University’s Archives Division. Self-assessment of computer literacy: Expert. Self-assessment of current document organization: Well organized. Level of difficulty in organizing documents: Easy. Main tasks:
- Follows up on the decisions taken by his/her supervisor and makes sure things get done.
- Manages his/her superior's agenda (fixes appointments according to priorities). - Manages the unit's accounts, registers expenses, produces reports, etc. - Receives and filters phone calls, gives out information. Welcomes visitors and
prepares necessary documents. - Sorts incoming mail according to priorities. Prepares responses and takes note of
any necessary follow-up. - Organizes meetings by sending out invitations, booking meeting rooms and
preparing the required documents. Takes part in meetings, takes notes, and ensures any necessary follow-up.
- Manages the local filing system and applies the required archival procedures. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents. - Participates in the planning of special events such as conferences or travels. - Accomplishes the tasks of a similar or lower position when needed.
194
47B157BParticipant #6: Francine
Age: 50-59 Years at the university: 16+ Position: Office Work Technician (OWT). Years in this position: 6-10 Use of the institutional scheme: No. Training in document management: - Self-assessment of computer literacy: - Self-assessment of current document organization: - Level of difficulty in organizing documents: - Main tasks: Francine was interviewed earlier than the others, as part of a previous pilot project. At the time, no survey was used to gather information before interviews. The information above was taken from the interview transcript, but a detailed task rundown was not available.
195
48B158BParticipant #7: Ginette
Age: 50-59 Years at the university: 6-10 Position: Student Files Technician (OWT). Years in this position: 1-5 Use of the institutional scheme: No. Training in document management: Received training from the University’s Archives
Division. Self-assessment of computer literacy: Expert. Self-assessment of current document organization: Very well organized. Level of difficulty in organizing documents: Easy. Main tasks:
- Participates in operations pertaining to the management of student files. - Follows up on the decisions taken by his/her supervisor and makes sure things get
done. - Updates the personnel files and prepares the appropriate documents for hirings,
nominations, promotions, vacations, absences, etc. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents. - Accomplishes the tasks of a similar or lower-ranked position when needed.
196
49B159BParticipant #8: Hortense
Age: 50-59 Years at the university: 16+ Position: Office Work Technician (OWT). Years in this position: 1-5 Use of the institutional scheme: No. Training in document management: Received training from the University’s Archives
Division. Self-assessment of computer literacy: Intermediate. Self-assessment of current document organization: Well organized. Level of difficulty in organizing documents: Easy. Main tasks:
- Participates in operations pertaining to the management of financial resources. - Participates in operations pertaining to the management of material resources. - Follows up on the decisions taken by his/her supervisor and makes sure things get
done. - Receives and filters phone calls, gives out information. Welcomes visitors and
prepares necessary documents. - Sorts incoming mail according to priorities. Prepares responses and takes note of
any necessary follow-up. - Manages the local filing system and applies the required archival procedures. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents. - Participates in the planning of special events such as conferences or travels. - Accomplishes the tasks of a similar or lower position when needed.
197
50B160BParticipant #9: Ingrid
Age: 20-29 Years at the university: 1-5 Position: Administration Technician (AT). Years in this position: 0-1 Use of the institutional scheme: No. Training in document management: In Cégep; also received training from the
University’s Archives Division. Self-assessment of computer literacy: Expert. Self-assessment of current document organization: Well organized. Level of difficulty in organizing documents: Easy. Main tasks:
- Participates in operations pertaining to the management of financial resources. - Participates in operations pertaining to the management of material resources. - Follows up on the decisions taken by his/her supervisor and makes sure things get
done. - Manages the unit's accounts, registers expenses, produces reports, etc.
198
51B161BParticipant #10: Jeanne
Age: 20-29 Years at the university: 1-5 Position: Office Work Technician (OWT). Years in this position: 1-5 Use of the institutional scheme: No. Training in document management: No. Self-assessment of computer literacy: Expert. Self-assessment of current document organization: Well organized. Level of difficulty in organizing documents: Easy. Main tasks:
- Participates in operations pertaining to the management of human resources. - Participates in operations pertaining to the management of financial resources. - Participates in operations pertaining to the management of material resources. - Follows up on the decisions taken by his/her supervisor and makes sure things get
done. - Manages his/her superior's agenda (fixes appointments according to priorities). - Manages the unit's accounts, registers expenses, produces reports, etc. - Receives and filters phone calls, gives out information. Welcomes visitors and
prepares necessary documents. - Sorts incoming mail according to priorities. Prepares responses and takes note of
any necessary follow-up. - Manages the local filing system and applies the required archival procedures. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents.
199
52B162BParticipant #11: Kim
Age: 40-49 Years at the university: 6-10 Position: Office Work Technician (OWT). Years in this position: 1-5 Use of the institutional scheme: Yes. Training in document management: Received training from the University’s Archives
Division. Self-assessment of computer literacy: Intermediate. Self-assessment of current document organization: Moderately well organized. Level of difficulty in organizing documents: Easy. Main tasks:
- Participates in operations pertaining to the management of human resources. - Participates in operations pertaining to the management of financial resources. - Participates in operations pertaining to the management of material resources. - Follows up on the decisions taken by his/her supervisor and makes sure things get
done. - Manages his/her superior's agenda (fixes appointments according to priorities). - Updates the personnel files and prepares the appropriate documents for hirings,
nominations, promotions, vacations, absences, etc. - Receives and filters phone calls, gives out information. Welcomes visitors and
prepares necessary documents. - Sorts incoming mail according to priorities. Prepares responses and takes note of
any necessary follow-up. - Manages the local filing system and applies the required archival procedures. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents.
200
53B163BParticipant #12: Lise
Age: 20-29 Years at the university: 1-5 Position: Office Work Technician (OWT). Years in this position: 0-1 Use of the institutional scheme: No. Training in document management: No. Self-assessment of computer literacy: Intermediate. Self-assessment of current document organization: Well organized. Level of difficulty in organizing documents: Easy. Main tasks:
- Participates in operations pertaining to the management of human resources. - Participates in operations pertaining to the management of material resources. - Follows up on the decisions taken by his/her supervisor and makes sure things get
done. - Manages the local filing system and applies the required archival procedures. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents. - Participates in the planning of special events such as conferences or travels.
201
54B164BParticipant #13: Manon
Age: 50-59 Years at the university: 16+ Position: Administration Technician (AT). Years in this position: 1-5 Use of the institutional scheme: Yes. Training in document management: Unfinished archival studies certificate; received
training from the University’s Archives Division. Self-assessment of computer literacy: Expert. Self-assessment of current document organization: Moderately well organized. Level of difficulty in organizing documents: Moderately easy. Main tasks:
- Participates in operations pertaining to the management of financial resources. - Manages the unit's accounts, registers expenses, produces reports, etc. - Accomplishes the tasks of a similar or lower-ranked position when needed.
202
55B165BParticipant #14: Nicole
Age: 40-49 Years at the university: 6-10 Position: Secretarial Agent (SA). Years in this position: 6-10 Use of the institutional scheme: No. Training in document management: No. Self-assessment of computer literacy: Intermediate. Self-assessment of current document organization: Well organized. Level of difficulty in organizing documents: Easy. Main tasks:
- Participates in operations pertaining to the management of human resources. - Participates in operations pertaining to the management of financial resources. - Participates in operations pertaining to the management of material resources. - Follows up on the decisions taken by his/her supervisor and makes sure things get
done. - Manages the unit's accounts, registers expenses, produces reports, etc. - Updates the personnel files and prepares the appropriate documents for hirings,
nominations, promotions, vacations, absences, etc. - Accomplishes the student files technician's tasks when needed. - Sorts incoming mail according to priorities. Prepares responses and takes note of
any necessary follow-up. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents. - Accomplishes the tasks of a similar or lower-ranked position when needed. - Manages his/her superior’s files.
203
56B166BParticipant #15: Olivia
Age: 20-29 Years at the university: 0-1 Position: Office Work Technician (OWT). Years in this position: 0-1 Use of the institutional scheme: Yes. Training in document management: In Cégep. Self-assessment of computer literacy: Intermediate. Self-assessment of current document organization: Moderately well organized. Level of difficulty in organizing documents: Moderately easy. Main tasks:
- Follows up on the decisions taken by his/her supervisor and makes sure things get done.
- Manages his/her superior's agenda (fixes appointments according to priorities). - Receives and filters phone calls, gives out information. Welcomes visitors and
prepares necessary documents. - Sorts incoming mail according to priorities. Prepares responses and takes note of
any necessary follow-up. - Organizes meetings by sending out invitations, booking meeting rooms and
preparing the required documents. Takes part in meetings, takes notes, and ensures any necessary follow-up.
- Manages the local filing system and applies the required archival procedures. - Takes notes and types various documents such as letters, reports, memos, etc.
Prepares page layout and verifies grammar and spelling. Signs certain documents. - Participates in the planning of special events such as conferences or travels.
