38
This article was downloaded by: [Towson University] On: 13 November 2014, At: 18:29 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK Human–Computer Interaction Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hhci20 Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones Jan o. Blom & Andrew F. Monk Published online: 12 Nov 2009. To cite this article: Jan o. Blom & Andrew F. Monk (2003) Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones, Human–Computer Interaction, 18:3, 193-228 To link to this article: http://dx.doi.org/10.1207/S15327051HCI1803_1 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.

Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

  • Upload
    andrew

  • View
    243

  • Download
    8

Embed Size (px)

Citation preview

Page 1: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

This article was downloaded by: [Towson University]On: 13 November 2014, At: 18:29Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,UK

Human–Computer InteractionPublication details, including instructions forauthors and subscription information:http://www.tandfonline.com/loi/hhci20

Theory of Personalizationof Appearance: Why UsersPersonalize Their PCs andMobile PhonesJan o. Blom & Andrew F. MonkPublished online: 12 Nov 2009.

To cite this article: Jan o. Blom & Andrew F. Monk (2003) Theory of Personalizationof Appearance: Why Users Personalize Their PCs and Mobile Phones, Human–ComputerInteraction, 18:3, 193-228

To link to this article: http://dx.doi.org/10.1207/S15327051HCI1803_1

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all theinformation (the “Content”) contained in the publications on our platform.However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness,or suitability for any purpose of the Content. Any opinions and viewsexpressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of theContent should not be relied upon and should be independently verified withprimary sources of information. Taylor and Francis shall not be liable for anylosses, actions, claims, proceedings, demands, costs, expenses, damages,and other liabilities whatsoever or howsoever caused arising directly orindirectly in connection with, in relation to or arising out of the use of theContent.

Page 2: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

This article may be used for research, teaching, and private study purposes.Any substantial or systematic reproduction, redistribution, reselling, loan,sub-licensing, systematic supply, or distribution in any form to anyone isexpressly forbidden. Terms & Conditions of access and use can be found athttp://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 3: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

Theory of Personalization ofAppearance: Why UsersPersonalize Their PCsand Mobile Phones

Jan O. Blom and Andrew F. MonkUniversity of York

ABSTRACT

Three linked qualitative studies were performed to investigate why peoplechoose to personalize the appearance of their PCs and mobile phones and whateffects personalization has on their subsequent perception of those devices. The1st study involved 35 frequent Internet users in a 2-stage procedure. In the 1stphase they were taught to personalize a commercial Web portal and then a rec-ommendation system, both of which they used in the subsequent few days. In the2nd phase they were allocated to 1 of 7 discussion groups to talk about their expe-riences with these 2 applications. Transcripts of the discussion groups were codedusing grounded theory analysis techniques to derive a theory of personalization ofappearance that identifies (a) user-dependent, system-dependent, and contextualdispositions; and (b) cognitive, social, and emotional effects.

The 2nd study concentrated on mobile phones and a different user group.Three groups of Finnish high school students discussed the personalization of

HUMAN-COMPUTER INTERACTION, 2003, Volume 18, pp. 193–228Copyright © 2003, Lawrence Erlbaum Associates, Inc.

Jan Blom recently completed his PhD on the psychological implications of per-sonalized user interfaces from the University of York; he is at Helsinki Institute forInformation Technology and Nokia Research Center. Andrew Monk investi-gates home use of information and communication technologies, mediated com-munication, and fun; he is a professor of Psychology at the University of York.

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 4: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

their mobile phones. Transcripts of these discussions were coded using the catego-ries derived from the 1st study and some small refinements were made to the the-ory in the light of what was said. Some additional categories were added; other-wise, the theory was supported. In addition, 3 independent coders, naive to thetheory, analyzed the transcripts of 1 discussion group each. A high degree ofagreement with the investigators’ coding was demonstrated. In the 3rd study, aheterogeneous sample of 8 people who used the Internet for leisure purposeswere visited in their homes. The degree to which they had personalized their PCs

194 BLOM AND MONK

CONTENTS

1. INTRODUCTION1.1. Defining Personalization1.2. The Research Methodology

2. STUDY 1: CONSTRUCTING THE THEORY2.1. Method

ParticipantsStage 1: The Training SessionStage 2: The Discussion Groups

2.2. ResultsDispositions to Personalize AppearanceEffects of Personalization on the User

3. STUDY 2: FINNISH MOBILE PHONE USERS3.1. Method

ParticipantsDiscussion Group Procedure

3.2. ResultsRefining the TheoryThe Generality of the ModelInter-Coder Agreement

4. STUDY 3: HOME VISITS4.1. Method

ParticipantsInterviews

4.2. ResultsRefining the TheoryPersonalization of the Home Computer

5. DISCUSSION AND CONCLUSIONS5.1. The Theory of Personalization of Appearance5.2. Relation to Previous Theories5.3. Reliability and Validity5.4. Epistemological Premises5.5. Design Implications

APPENDIX A: STUDY 1 FOCUS GROUP PROTOCOLAPPENDIX B: THEORY OF PERSONALIZATION CATEGORIES

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 5: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

was found to be well predicted by the dispositions in the theory. Design implica-tions of the theory are discussed.

1. INTRODUCTION

People commonly personalize places and objects in their surroundingsand there is a research literature describing how people personalize their of-fice spaces (Scheiberg, 1990; Wells, 2000), isolated environments such aspolar stations (Carrere & Evans, 1994), dormitory rooms (Vinsel, Brown,Altman, & Foss, 1980), and hospital wards (Holahan & Saegert, 1973). Thesame phenomenon can be observed in the way people decorate certainelectronic products. Mobile phones are now sold with replaceable colorcovers. Operator logos and ringing tones can be downloaded from theInternet. Extensive changes can be made to the appearance of the desktopof a PC as well. Web portals, such as Yahoo! and Excite, allow the user tochange the style scheme of the page. Some manufacturers have taken morenotice of this trend than others have. Each edition of Microsoft Windowshas introduced new decorative features for personalizing the desktop.Nokia phones demonstrate a similar developmental trend. The top of therange model currently has 10 features that can be used to personalize thevisual appearance and sound of the phone. This article presents a qualita-tive study of why people personalize and how this process changes theirview of the product concerned.

1.1. Defining Personalization

The concept of personalization broadens in moving from the personaliza-tion of the environment to personalization of technology. Wells (2000) de-fined the former as “the deliberate decoration or modification of the environ-ment.” Blom (2000) defined the latter “as a process that changes thefunctionality, interface, information content, or distinctiveness of a system toincrease its personal relevance to an individual.” An important difference isthat Blom’s definition includes changes to what the product can do as well asits physical appearance.

The context of information and communication technology is also ratherdifferent from the context of an office space or a dorm in that personalizationcan be initialized by the device as well as by the user. Blom (2000) pointed outthat whether personalization is system- or user-initiated can be viewed as a di-mension rather than as a dichotomy as often both the system and the user par-

THEORY OF PERSONALIZATION OF APPEARANCE 195

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 6: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

ticipate in the process. For instance, the system may recognize that some as-pects of the user’s interaction with the software could be optimized but thechange does not take place until the user has authorized it. Blom and Monk(2001) distinguished between personalization and personification of Webpages: The latter refers to the image presented by a Web page, whether itpresents itself as human or machine like. They found that the effects of per-sonalization on subjective ratings of workload and engagement can dependon how the system presents itself.

Most of the current academic interest in the personalization of technol-ogy focuses either on classifying various personalization methods (cf.Kobsa, Koenemann, & Pohl, 1999) or on describing the underlying market-ing principles (Peppers & Rogers, 1997). The former could be described asa system-centered approach, and the latter as a business-oriented perspec-tive. This article sets out a user-centered account of personalization with theaim of informing the design of information and communication technologyproducts.

1.2. The Research Methodology

Blom (2000) defined a taxonomy of the potential motivations a user mayhave for personalization. This distinguished between work-related and so-cially related motivations. Work-related motivations are concerned with ac-commodating work-related goals. For example, an operating system might al-low you to change the order or content of menus or increase all font sizes toaccommodate visual impairment. An example of system initiated personal-ization with work-related motivation would be a recommendation system(Guttman, Moukas, & Maes, 1998) that suggests a book that the user might en-joy on the basis of previous purchases. Socially related motivations have to dowith eliciting emotional responses in the personalizer (Scheiberg, 1990) andexpressing identity (Goffman, 1959) to oneself and others. Changing thecover on a mobile phone, for example, has no effect on the functionality ofthe phone but is nevertheless valued by users.

Blom’s (2000) taxonomy is speculative and limited. The purpose of thethree studies presented here is to provide a new account of the user experi-ence associated with personalization of appearance, based on empirical data.As the aim was to discover a new theory rather than verify existing knowl-edge, the nature of the approach was qualitative. Discussion groups were runand the transcripts were analyzed using grounded theory techniques, whichimply a “general methodology of analysis linked with data collection that usessystematically applied set of methods to create an inductive theory” (Glaser,1992, p. 16). Thus, in grounded theory analysis (GTA), “a researcher does not

196 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 7: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

begin a project with a preconceived theory in mind. Rather, the researcherbegins an area of study and allows the theory to emerge from the data”(Strauss & Corbin, 1990, p. 12).

Of course, given the background described earlier, we cannot pretend thatthe research process was not guided by previously existing knowledge. Nev-ertheless, grounded theory was selected to ensure that data collection, analy-sis, and eventual theory stand in close relation to one another (see Strauss &Corbin, 1990). The issue of to what extent previous knowledge influenced uswill be taken up in the discussion.

The data to be presented in the following sections come from discussiongroups in which the focus was the personalization of PCs and mobile phones.There were three studies. The first study provides the major theoretical im-port of the article and will be discussed in some detail. It involved a two-stageprocess in which users were individually introduced to examples of Web per-sonalization and then later brought together in seven discussion groups. Thesecond study was a grounded theory analysis of three discussion groups con-sisting of female Finnish teenagers talking about their mobile phones. Thisserves to confirm the generality of the theory generated in the first study witha rather different user group. In the third study, eight individuals were visitedin their homes to view and discuss what they had done to personalize theirPCs. This study also provides some tentative confirmation of elements in thetheory.

In the first study, the discussion groups focused on two kinds of personal-ization: personalization of physical appearance, and collaborative filtering(cf. Guttman et al., 1998; Blom, in press). Separate GTAs of these two person-alization facilities were performed and they turned out to be rather different.Only the analysis of personalization of appearance is described here.

2. STUDY 1: CONSTRUCTING THE THEORY

The participants for this study were university-based Internet users. ThePCs available to this population tend to be for shared use so that personal-ization of the wallpaper or desktop is not an option. However, many Webportals provide personalization features that are available to these users. Auser can register with the portal and record certain personalized settingssuch as color schemes, welcome messages, and categories of news items.Preparatory research revealed that this form of personalization was notcommon, even among this population of relatively frequent Internet users.A two-stage method was adopted in which the participants were individu-ally instructed in the use of personalization before the discussion groupswere held.

THEORY OF PERSONALIZATION OF APPEARANCE 197

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 8: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

2.1. Method

Participants

An opportunity sample of 21 women and 14 men was recruited from thepopulation of undergraduate and postgraduate students at the University ofYork. The age of the participants varied between 18 and 36, with the meanage being 21. The participants received either a payment or a course credit forparticipating in the study. To avoid awkward group dynamics, separate maleand female interview groups were formed. Note that comparisons are notmade here between the male and female groups. It was not considered safe tomake generalizations regarding individual differences without a larger sam-ple size.

As the questions were centered on Web-based personalization, the userswere selected to be relatively frequent Internet users, doing leisure-basedbrowsing at least once a week (most of them a few times per week). The extentof IT usage was assessed with the help of a short questionnaire at the end ofthe second session. The first author was the facilitator for the groups. He was aPhD student at the time and did not know any of the participants. The facilita-tor took care to present himself as an equal and not to impose ideas or confor-mity within the groups. Similarly, to minimize the convergence of ideascaused by social conformity (Asch, 1951), individuals studying a certain sub-ject, and hence potentially knowing each other, were allocated across thegroups.

Stage 1: The Training Session

The aim of the training session was to acquaint the participants with thetwo services that were to be discussed in the groups. Each participant wasgiven individual tuition in a session that lasted approximately 25 min. Eachparticipant first became a member of MyExcite. They were then asked to per-sonalize the appearance of the main page by changing the welcome slogan,an image displayed on top of the page, and the color scheme of the page (seeFigure 1). To further exercise the use of this site, an e-mail was sent to the uni-versity e-mail inbox of the participants a few days prior to the discussiongroup asking them to check their e-mail inbox on Excite.

Stage 2: The Discussion Groups

The number of participants in the groups varied between three and six,with the average being five, and the sessions lasted approximately 90 min.The questions that were asked were open ended and designed to gradually fo-

198 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 9: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

cus on the implications of personalization. A list of pre-determined,open-ended questions was used to ensure the same questions would be askedin the groups (see Appendix A). The first author acted as the facilitator in eachof these two stages.

2.2. Results

GTA was performed on the transcribed conversations by the first author.The GTA method that was adopted was based on that of Strauss andCorbin (1990), with the exception that the paradigm model, which involveslinking subcategories to categories using “conditions,” “context,” “action–interactional strategies,” and “consequences” was not utilized. We agreewith Glaser (1992), who suggested that resorting to the paradigm modelwould involve preconception and forcing theoretical concepts on data. In-stead, argued Glaser, the analyst simply codes for categories and propertiesand lets conceptual relationships emerge from the data.

The GTA took place as follows. First the extracts were characterized forcontent by open coding. All the utterances relevant to the process of personal-ization were labeled. This stage resulted in the identification of 51 codes relat-ing to the personalization of appearance. Axial coding was then performedwhere the codes were grouped into a smaller set of categories and connec-tions were developed between categories and their subcategories. The 51 ini-tial codes were reduced to a set of 34 categories. Eighteen of these were identi-

THEORY OF PERSONALIZATION OF APPEARANCE 199

Figure 1. The participants were familiarized with Excite, where one is able to choose the(a) welcome slogan, (b) color scheme, and (c) the image that is displayed on top of thepage.

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 10: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

fied as main categories, and 16 as subcategories. To keep the article to amanageable length, only the main categories are discussed here. Appendix Blists the main categories with examples of the comments that exemplify them.Figure 2 illustrates the nature of this list with the categories frequency of use ofsystem, feeling in control, and improved aesthetics as examples. The first col-umn gives the name of the category devised by the coder. The next two col-umns give the total number of instances of utterances instantiating this cate-gory and the number of groups (out of seven) that had at least oneinstantiation, respectively. The last column gives a typical example of an ut-terance illustrating the category.

The final stage in the analysis was to conduct selective coding. This is theprocess of integrating and refining the theory (Strauss & Corbin, 1990). Thecore concept here, around which the opinions of the participants folded, is theprocess of personalization. The categories are divided into those described byparticipants as being relevant to initiating the personalization process (dispo-sitions) and those described as effects of personalization. Figure 3 details thistheory of personalization, the super-ordinate categories being printed in bold.The dispositions were divided into user-, system-, and context-relatedsuper-ordinate categories, the effects into cognitive, social, and emotionalsuper-ordinate categories. The dispositions and the effects are discussed inmore detail in the following.

Dispositions to Personalize Appearance

As shown in the left column of Figure 3, the dispositional categories weredivided to user-dependent, system-dependent, and context-dependentsuper-ordinate categories. The individual dispositions are discussed in the fol-lowing.

Frequency of Use of System. When asked which device one would wishto personalize, the most popular were mobile phones, portals, e-mail applica-tions, and desktops of PCs. What is common to all of these is the high fre-quency of usage. Participants indicated that, if a device is used frequently, thelikelihood of personalizing it increases:

I would personalize my desktop. Just because it’s the one thing you lookat, every time you turn on the computer, first thing you ever see.

Ownership of System. When prompted why so many have personalizedtheir personal computers but are nevertheless not that interested in being ableto personalize Excite, a typical response was

200 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 11: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

I think the computer is mine. And nobody else touches it. Well oh, boy-friend I guess, but nobody’s allowed to change anything because that’sthe way I’ve put it.

In contrast,

on Excite it’s just someone else’s control of what’s the content.

Knowledge of Personalization. Most participants were not aware thatthey could change the appearance of a portal. Many expressed the view that

THEORY OF PERSONALIZATION OF APPEARANCE 201

Category InstancesGroups(Total 7) Sample Extract

Frequency ofuse of system

22 6 I would personalize my e-mail because you see itso often.

Feeling incontrol

8 4 [Personalization] Makes you feel more in controlso you actually know what you’re doing.

Improvedaesthetics

9 3 It’s good that aesthetically you can change it[Excite] for what you like personally.

Figure 2. Illustration of the table of categories that is displayed in Appendix B. Alto-gether 34 categories arose as a result of the GTA.

Dispositions to PersonalizeAppearance

Effects of Personalization ofAppearance on User

User CognitiveFrequency of use of S Ease of useOwnership of S Improved aestheticsKnowledge of P Recognition of S

System SocialEase of P Reflection of personal identityEffectiveness of P features Reflection of group identity

Context EmotionalSocioemotional context of use (SEC) Familiarity with S

S feels personalFeeling in controlFeeling of ownershipRelease from boredomFunPositive associations

Note. S = system or product; P = personalization.

Figure 3. An initial theory of personalization. The categories that arose as a result of theGTA in the two-stage study. The theory consists of dispositions and effects.

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 12: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

awareness of the ability to personalize and knowledge of how to personalizewould result in an increased tendency to achieve this process:

I didn’t realize before that you could change [personalize the appear-ance]. I thought it’s on there and someone’s designed the Web site andthat’s it really. Now that I know it’s possible I’ll probably look into it.

Ease of Personalization. The other side of the coin to personalizationknowledge is system difficulty. At least to novice users of a given device, thelevel of perceived system difficulty can affect an individual’s disposition topersonalize. When prompted what devices the participants would like to per-sonalize, one of the discussants answered as follows:

Just things on my mobile phone and my laptop. I like doing it becausethey are all easy to change.

Effectiveness of Personalization Features. It was found that the person-alization features have to be effective in terms of their ability to elicit emo-tions, express the identity of the user, or evoke cognitive effects such as ease ofuse. On Web sites it is often the case that the user does not have much choicewhen making changes. That is, the personalization features are not effective:They do not correspond to the needs and preferences of the user.

And with Excite you could just go for the color schemes that they hadset, you couldn’t like say change the background for this and then theheading title to this. It was their scheme so you chose one of theirschemes.

The lack of relevant material or options may be one of the greatest barriers touser-initiated personalization. Needless to say, this situation can be avoidedby concentrating on designing a wide range of personalization items and test-ing them with users to ensure they are effective.

Socioemotional Context of Use. The degree to which the use of a systemwas associated with a socioemotional context of use (SEC) was found to influ-ence users’ dispositions to personalize. Thus, SEC becomes a relevant cate-gory when the device enables the use of features such as Short Message Ser-vice (SMS) or e-mail. To illustrate this, one of the participants preferred thepersonalization of e-mail to that of a word-processing application. Whenasked why, she answered,

202 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 13: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

On Word when you sit down and write an essay or something and youdon’t want those colorful flashy things distracting you. But with e-mailit’s not just work. You’re writing to friends and things.

Other possible socioemotional usage functions would naturally includegames or surfing. Note that this category was included in the super-categorycontext rather than system because the user’s motivation also plays a rolehere. In other words, even if a system includes leisure-based functions, an in-dividual is not necessarily motivated to use these. Thus, both socioemotionalfunctionality and usage motivation are requirements for socioemotional us-age context.

Users who exhibited a task-based motivation were more likely to personal-ize the functionality or information content of a given system as these aremore directly related to accomplishing a task.

I don’t care if it’s blue or if it’s green or whatever [the Web site]. It’s justreally the information that I want.

If a task-based personalizer decides to change the appearance of a product,this is likely to evoke cognitive effects, for instance make the system more rec-ognizable:

I changed my mobile’s ring and volume but other than that … it’s a mo-bile, it’s there for a purpose.

Note that an individual who has previously personalized the appearance ofa device for task-based reasons could at another time personalize forsocioemotional reasons. This would be likely to take place in spare time:

But it’s [personalization of appearance] just when you’re time-wastingreally. Something to do.

Effects of Personalization on the User

The previous section detailed the dispositions to personalize described bythe participants. This section discusses what they described as the effects ofpersonalization (see Figure 3 for overview).

Ease of Use. The prototypical cognitive effect is perceived ease of use. Itmight seem odd that a nonfunctional change such as personalization of ap-pearance, can affect this perception. There are, however, attributes related tothis feeling that can be accommodated by making changes to the appearance.

THEORY OF PERSONALIZATION OF APPEARANCE 203

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 14: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

The perception of ease of use was for instance often mentioned in conjunctionwith the system becoming more aesthetically pleasing or making parts of thesystem more recognizable with the help of personalization:

Why I like it personalized? … you can make it easier to use, you can seethe buttons.

Improved Aesthetics. An obvious, and often natural, consequence ofchanging the appearance of a system is that it becomes more aestheticallypleasing:

Let’s say it’s a Web site that my friends are using. I mean there are somecolors that you just don’t find attractive. So this way it does not matterwhat the initial color is.

The reason for not casting this category as an emotional effect is becausethe users often associated the aesthetic aspects of the system with task accom-modation. Improved aesthetics would thus involve making the system “easierto the eye,” which is perhaps more readily characterized as a cognitive ratherthan as an emotional effect:

I changed Microsoft Word, the toolbar, the fonts, type, size, and color.Mainly because I was working on it all day everyday and it was easieron my eye.

Recognition of System. It is often the case, especially with mobilephones, that several people within a peer group own the same model. Thiswould consequently lead to a need to personalize the phone:

Jo has got one [Nokia 3210] and my other house mate has one so there’sthree same phones all sitting in the room and it’s like “which one ismine.”

As with improved aesthetics, also recognition of system could be con-ceived of as an emotional effect. The main implication of this category, how-ever, was a functional one: Personalization made the device distinctive thusenabling the users to gain access to their devices, which in turn makes this ef-fect inherently cognitive.

Reflection of Personal Identity. Personalization was often compared tothe use of clothes as a way to distinguish oneself from others:

204 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 15: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

• It’s the same reason not everybody wears the same clothes. They want tobe distinctive and different … so it’s just another extension, electrical de-vices you have with you all the time.

• You did that in school actually with school uniform.• School uniform?• Fold up your shirt.• High heels, low heels, trainers.• Short sleeves, long sleeves, collars up collars down.

Identity is thus more than recognition. One is distinguishing oneself fromothers and at the same time identifying with some subgroup or value set:

• It’s a message to other people though [a personalized phone].• It’s something about your personality.• Yeah this is me, this is what I like, this is what I don’t like.

Reflection of Group Identity. Personalization could also take place to ex-press or improve aspects of group identity. Consequently, personalizationsometimes involved a number of individuals. For instance in the utterance fol-lowing, a participant’s friend had personalized her desktop. The graphiccould then create a mutual, shared concept, a factor binding these persons to-gether:

You always tend to leave it [personalized desktop] if one of your mateshas done it. Then every time you turn your computer back on … itmakes you smile because you know your friend has done it.

Familiarity With System. The fact that personalization was found to re-sult in a feeling of familiarity is not surprising—the changes that are made bythe user imply that the user focuses attention on various features of the sys-tem. It is expected that this attention results in a higher degree of familiaritywith the system. Several ways were used to describe this feeling. For instance,it was noted that personalizing a Web site enables one to “instinctively knowwhat this thing is.” The personalized site could also act as a “familiar startingpoint” from which to start traversing to other sites.

System Feels Personal. Perhaps unsurprisingly, participants often com-mented that personalization made a product feel more personal:

I liked that [personalization]. Your own little greeting that made it a bitmore personal.

THEORY OF PERSONALIZATION OF APPEARANCE 205

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 16: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

It is possible that this is just a reflection of the prompt to describe personaliza-tion and in a sense this whole section is an attempt to define the various mean-ings “personal” might have. Nevertheless it is included here as participantsmentioned it so frequently.

The term personal was often used in conjunction with feeling in control or afeeling of ownership, included as separate categories in the following.

It kind of makes you feel more like it’s yours. It’s actually personal toyou.

Feeling in Control. The data led to the conclusion that personalizing theappearance of a system can increase the feeling of control over the device:

If it’s your own desktop you’ve got control over it and no one else ischanging it whereas on Excite it’s just someone else’s control of what’sthe content.

Feeling of Ownership. The importance of ownership is highlighted bythe following extract in which personalization of a device was compared tothe process of decorating one’s room:

It makes it yours [personalization]. It’s like on campus, in your room,you stick up posters up, don’t you. And you stick things you want onyour wall. Put your own duvet cover on.

Release From Boredom. Users sometimes get bored with aspects of theappearance of the system and want to change them. The personalized stimulihave ceased to be effective in eliciting emotions in the user:

On one hand it’s nice to have it personalized so it’s yours so you recog-nize it. But it’s also nice to change it every so often. Because you getbored with it being the same all the time. You want something new.

This category would explain the need to personalize something again andagain. As time goes by the user becomes desensitized to the previous person-alization, resulting in the need to re-personalize the device.

Fun. The comments that were concerned with reacting to the personal-ized device often referred to feelings of amusement or fun:

I have a U2 desktop theme on my computer (laughing). It plays littletunes to me. It’s quite amusing really.

206 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 17: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

Associating feelings of fun with the use of a personalized system is furtherindicated by the fact that words such as toy, gimmick, and gadget were used todescribe these personalized devices. Statements consistent with this are,

• It’s fun, isn’t it [personalization of appearance].• Yeah, it’s just something to play around with.

Positive Associations. Several comments were made that referred to per-sonalization evoking positive associations in the user. Sometimes, generalpositive items were used to elicit these feelings:

I’m sticking to the flower, because I turn it on and think, oh it’s sunnyoutside.

Someindividuals,on theotherhand,used items thatwerespecific to their lives:

I put the mountains [on the desktop of a PC] because I love skiing andbecause it reminds me of my house. So each time I put it on, I was like:“Oh yeah.”

This section has been concerned with describing the GT analysis resultsthat led to the emergence of dispositions and effects of personalizing the ap-pearance of a product. The next section discusses the first follow-up study.

3. STUDY 2: FINNISH MOBILE PHONE USERS

To examine the generality of the model described in Section 2, discussiongroups were run with a different user population and a different device. Thefocus was mobile phone personalization by Finnish teenagers. This combina-tion of device and user population was chosen as one in which there was ahigh degree of personalization and in which the individuals were differentfrom the UK students used in the two-stage study. Overlap between the cate-gories generated here and in the previous study suggests generality. Wherenew categories occur the theory may be refined to include them. The studyalso looks at agreement between coders.

3.1. Method

Participants

Each of the three discussion groups consisted of five female high schoolstudents. The participants were students at a Finnish high school in Helsinki.

THEORY OF PERSONALIZATION OF APPEARANCE 207

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 18: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

The age of the participants varied between 16 and 18, with the mean age be-ing 16.5 years. Opportunity sampling was used to recruit the participants andthey received a payment for taking part in the study. To be able to participate,the participant had to own a mobile phone. Consistent with Study 1, precau-tionary measures were taken to minimize potential biases arising from the so-cial dynamics of group settings. First, none of the participants were personalacquaintances of the facilitator. Second, the participants within the groupswere also relatively unfamiliar with each other as individuals from any givenclass were allocated across the groups.

Discussion Group Procedure

The three discussion groups were run in a small-size classroom of the highschool the students were attending. The interviews took place in the after-noon and each session lasted about 75 min. The initial stage of the sessionconsisted of the participants introducing themselves and explaining their mo-bile phone usage history and patterns. Toward the end, the questions startedfocusing on the motivations and effects of personalization. The sessions werevideotaped and later transcribed. Note that the interviews took place in Finn-ish. The extracts that are displayed below are therefore translations.

3.2. Results

Refining the Theory

As in Study 1, the first author acted as both the facilitator of the sessionsand as the coder and analyst of the GTA. There were altogether 97 utterancesrelevant to personalization, out of a total of 219, that could not be assigned tothe existing categories or that called for elaborating a previously existing cate-gory. A further analysis on these resulted in the construction of six new cate-gories: cost of personalization, absence of technical constraints, seasonal andmedia influences, peer influence, attachment to system, and accommodatingcurrent emotional state. The first four are dispositions, whereas the last twoare effects of personalization. In addition to adding new categories, the cate-gory reflection of group identity was further elaborated.

Cost of Personalization. It is easy to understand that the cost of personal-izing can in many cases affect a user’s disposition to personalize. This is espe-cially true for users from younger age groups.

I suppose when these things [personalization items] get cheaper, then Icould consider personalizing the phone.

208 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 19: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

Absense of Technical Constraints. The phone models of a few users werequite old. Consequently, these phones were not associated with as many per-sonalization options as some of the more recent models. This sometimes ledto a situation in which the need to personalize existed but, due to the technicalconstraints, the user would not be able to accomplish this process:

I’ve noticed that the 5110 [a phone model], which you cannot changethat much, people often say that they would change it more if theycould.

Seasonal and Media Influences. Different kinds of seasonal and mediathemes relating to festivals, time of the year, or currently popular cartooncharacters were reflected in the personalization of phones. As the interviewstook place shortly after Christmas, one of the participants referred to using a“jingle bell” theme:

Quite a few people around Christmas time order “jingle bell” stuff totheir phones. And when there was this Star Wars boom, I was really ex-cited as well. I ordered the “Darth Vader” theme tune.

Peer Influence. Personalization was being incorporated by many of thefriends of the participants and this was a major cause of personalization. Thiscategory was cast as a contextual disposition.

When I changed the logo for the first time, one of my friends had thesame kind of a thing on her phone. I’m not that bothered, but I supposeI will go on and order it as well. You get influences from your environ-ment.

Attachment to System. One of the participants referred to her phone as a“living thing, a pal.” Another one regarded it as a “personal friend.” A thirdperson had actually named her phone. Someone even assigned it the status ofa pet. These users had developed a feeling of attachment toward the device.Personalization was a process contributing to this:

Well, you get attached to the phone and it’s more comfortable, as com-pared to if you could not personalize it.

Accommodating Current Emotional State. Several members of the dis-cussion groups personalized to accommodate their current emotional states.Personalization could thus be seen as a way of regulating one’s current feel-ings. As compared to identity-related personalization, changing the device as

THEORY OF PERSONALIZATION OF APPEARANCE 209

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 20: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

a function of the current emotion state refers to a more transient form of per-sonalization:

I change the ringing tone constantly, just according to how I’m feeling.At the moment I’ve got Red Hot Chili Peppers’ “Other Side.”

Elaborating the Category Reflection of Group Identity. There was awealth of comments relating to the previously established category, reflectionof group identity. The large amount of data associated with this category re-sulted in further refining it. First, a need to express aspects of the group identitywas reported. The supposition here is that personalization can act as a mecha-nism for creating and expressing mutual values. In the following extract, theaspect of common identity is interest in cars:

This weekend we were spending some time together and then we got anidea to, you know 19–20 years old people like cars, so we got an idea togo and order a car logo to a mate’s phone. So we had to find a logo to hisphone, a Ford one or something. We browsed through the net and triedto find as nice a Ford as possible.

Comments relating to the identity expression were often associated withimproving the group cohesion (Schachter, Ellertson, McBride, & Gregory, 1951)by incorporating personalization as a shared activity. Finally, personalizationalso acted as a way of having fun with peers. The strong group membership ele-ment associated with mobile telephones is supported by similar findings re-garding the mobile phone SMS culture. Kasesniemi and Rautiainen (2002)reported on the collective behavior that Finnish teenagers’ text messaging in-corporates. Text messages are circulated among friends, composed together,and read together. Thus, through the use of SMS, conclude these authors,teenagers collectively engage with each other and sustain their teenage rela-tions.

The Generality of the Model

Of the 18 main categories identified in Study 1, 14 also occur in this study.Figure 4 illustrates the overlap. The categories present in both sets of datahave not been changed. The new ones are underlined and the categories thatonly occurred in Study 1 are shown in italics.

The high degree of overlap shows that the theory stands up well when ap-plied to the Finnish mobile phone users. Six new categories were added. Ofthese, the system dispositions cost of personalization and absence of technicalconstraints can be thought of as extremes of ease of personalization, which

210 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 21: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

did not occur here. Accommodating current feeling is a new emotional effectand the two environmental dispositions seasonal and media influences andpeer influence were not apparent in the earlier discussion groups using Eng-lish university students.

Whereas the new categories (underlined in Figure 4) would appear to beexplicable in terms of the different priorities of the users, the missing catego-ries (italicized in Figure 4) may be explained by differences in the devices dis-cussed. In Study 1 the focus was PC and mobile phone personalization. InStudy 2 it was only mobile phones. The categories that did not occur in thelatter study were ease of personalization, ease of use, familiarity with system,and feeling in control, categories that it may be hypothesized are more closelyrelated to PC usage rather than mobile phone use for this population.

Inter-Coder Agreement

As a further test of the model presented in Figure 4, three Finnish-speakingjudges, naive to the coding categories by the authors, were each given rudi-mentary instructions and a transcript of one of the Finnish groups. The codesgenerated were then compared with those of the investigators to assessinter-coder agreement.

THEORY OF PERSONALIZATION OF APPEARANCE 211

Dispositions to PersonalizeAppearance

Effects of Personalization ofAppearance

User CognitiveFrequency of use Ease of useOwnership of S Improved aestheticsKnowledge of P Recognition of system

System SocialEase of P Reflection of personal identityEffectiveness of P features Reflection of group identityCost of P EmotionalAbsence of technical constraints Familiarity with S

Context S feels personalSocioemotional context of use (SEC) Feeling in controlSeasonal and media influences Feeling of ownershipPeer influence Release from boredom

FunPositive associationsAttachment to systemAccommodating current emotional state

Note. S = system or product; P = personalization. The new categories are underlined. The cate-gories that only occurred in Study 1 are shown in italics.

Figure 4. Coding overlap between the model developed from Study 1 and Study 2.

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 22: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

The coding process for these external coders was as follows. Havingviewed a video of the discussion group they were asked to mark any causesand effects of personalization mentioned in the transcript. In the second ses-sion, the first author asked the coder to talk through the causes and effects.These were then cut out and pasted onto cards for the third and final session.In this session the coders sorted these cards into categories and, where appro-priate, grouped the categories under super-categories, thus producing a con-ceptual model of personalization.

Goodman and Kruskal’s tau was used as a measure of association to deter-mine the level of agreement between the coding of the authors and the exter-nal coder. This assesses how well the items that are put together by one coderpredict the items that are put together by the other. The items in this analysisare the utterances coded by the authors. As one might expect given the shorttime available to them, the external judges did not categorize as many ex-tracts as the author had done. A category “not coded by the external coder”was thus created. Goodman and Kruskal’s tau was computed from the result-ing cross-tabulation using the categories of the authors as the independentvariable and the external coders’ categories as the dependent variable. Twoindexes were calculated, one for all the utterances and one excluding the ut-terances not coded by the external coder. Tau, which has a minimum of 0 anda maximum of 1, was .55, .54, and .58 for groups 1, 2, and 3, respectively,when the noncoded instances are included. All of these were significant at the.001 level. When excluding the noncoded extracts, the respective taus rose to.84, .86, and .80, indicating a very high degree of agreement.

Goodman and Kruskal’s tau makes no assumption about the equivalenceof the categories used, only that the same items go together. For instance, ifcoder A referred to a set of utterances as happiness and coder B consistentlyreferred to the same set of utterances as sadness they could still get a tau of 1,indicating a high level of agreement. To rule out this possibility, the semanticsimilarity between pairs of categories was assessed by the first author. Of a to-tal of 33 pairs, there were 18 clear semantic matches, 14 reasonably goodmatches and one pair where there was no similarity. Overall these data indi-cate a high level of spontaneous inter-coder agreement that validates our in-terpretation of the transcripts.

4. STUDY 3: HOME VISITS

In a further attempt to widen the examples of personalization examined,the third qualitative study consisted of home visits to frequent Internet users.A professional recruitment agency was asked to find eight people, with heter-ogeneous backgrounds, who used PCs at home for leisure purposes. Thesepeople were then visited and asked to demonstrate how they had personal-

212 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 23: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

ized their machines. The study thus (a) widens the user populations sampledand (b) allows us to see what personalization people had actually done, ratherthan simply record what they said they had done.

4.1. Method

Participants

Eight PC users were interviewed in their own homes. The participants wereall living inYork,and theywereacquired througharecruitingagency.Thecrite-rion given to the agency was that participants were using the Internet at homefor an average of at least 2 hr a week, with at least 25% of this usage being e-com-merce and surfing related rather than e-mail or chat. The relatively high leisureusage requirementwas toensure that theparticipantswouldbeable toproperlyreflect on Web-based personalization. The sample was older and had more var-ied occupations than in the other studies (see Figure 5).

Interviews

The interviews lasted about 90 min each and were conducted in the homesof the participants. The authors jointly facilitated the sessions. They startedwith open-ended general questions about who in the home used the com-puter. The participants were asked to demonstrate the ways in which theirPCs had been personalized and the interviewers demonstrated personaliza-tion features on the Yahoo! Web portal to generate further discussion. Finally,the participant was asked to show the interviewers their mobile phone andthese personalization features were discussed. At the end of the sessions, theparticipants filled in a short questionnaire on computer use and personaliza-tion.

4.2. Results

Refining the Theory

The model of personalization accounted for most of the content of the in-terviews. There were only six extracts that were relevant to the personaliza-tion process that could not be accounted for. One category was created to ex-plain these extracts: new to system. This category was included in thedispositional super-category user. Half of the participants stated that they hadbeen personalizing their machines especially when it was new. This waspartly because it would be common to regard personalization as a noveltywhen having acquired the computer:

THEORY OF PERSONALIZATION OF APPEARANCE 213

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 24: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

First thing it was a novelty [personalization] so I kept doing it every cou-ple of weeks.

This initial personalization also provides the user with a chance to explore anovel system.

When I first got the computer, I used to fiddle about with the appear-ance all the time … Get acquainted a bit.

Someone might argue that the kind of dynamics this category implies threat-ens the rest of the model—perhaps personalization is only relevant when theproduct is new. However, this was not the case. The participants that personal-ized accomplished this process at least occasionally though they had pro-gressed beyond the initial usage stage. The category new to system thus simplyidentifies a point in time when one is especially likely to personalize.

Personalization of the Home Computer

There was considerable variation in the extent to which the participantshad personalized their PCs. A few had never personalized their machines,

214 BLOM AND MONK

Participant Age OccupationMain Area ofComputer Use

Main Area ofWeb Use

Shopping Viathe Web

H. L. (F) 25 – 34 Student Wordprocessing

Work Clothes, games,CDs, books

R. T (M) 45 – 54 Manager Surfing andgames

N/A N/A

J. Ma (M) 18 – 24 Processoperative

Chat andgames

Finance,entertainment

Has not doneshopping

W. H. (F) 25 – 34 Accountsmanager

E-mail andsurfing

Shopping Clothes,jewellery,CDs, books,pictures

T. S. (M) 45 – 54 Driver E-mail Finance Books, CDs,clothes

J. M. (F) 35 – 44 Labtechnician

Surfing Entertainment Books, CDs,holidays

S. R. (M) 45 – 54 Manager Surfing Entertainment Tickets,holidays, PCparts

I. C. (M) 45 – 54 Teacher Surfing andwordprocessing

Entertainment CDs, holidays,wine, shoes,hotel rooms

Note. Participant’s sex is in parenthesis.

Figure 5. Participant details in Study 3.

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 25: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

whereas others were relatively frequent personalizers having, for example,book-marked Web sites from which they would download desktop themes.Figure 6 illustrates the frequency and extent to which each participant hadpersonalized their computer. It also shows the extent to which the participantsfulfilled the dispositional categories knowledge of personalization, SEC, andownership of system. That is, it shows how the participants vary in terms oftheir knowledge to personalize, usage context (whether task-based orsocioemotional), and ownership of the system.

In the figure, three distinct groups can be identified as a function of the fre-quency and extent of personalization. S. R. and R. T. are frequentpersonalizers and they utilize the whole range of personalization features. J.Ma, H. L., and W. H. are occasional personalizers and only one of them (H.L.) utilizes effectively the range of the possible personalization features.Finally, J. M., I. C., and T. S. have never personalized their PCs.

The notion of dispositions leading to effects used to structure the modelsuggests that it should be possible to predict the frequency and extent of per-sonalization from the next three columns in Figure 6. In line with the model,the participants with a high scope of personalization score high on thedispositional categories knowledge of personalization, SEC, and ownershipof system. Moreover, these values tend to decrease for the users in the me-dium and low personalization groups.

To further illustrate this trend, a dispositional index is displayed in the lastcolumn of Figure 6. This is the average of the previous three columns thatdemonstrate knowledge of personalization, SEC, and ownership of system.As can be seen, the index predicts well the frequency and extent of personal-ization for each of the groups thus providing additional support to the theory.S. R. and R. T., the users from the high personalization group, score higheston the dispositional index; J. M., I. C., and T. S., the ones who had never per-sonalized, score lowest on the index. Furthermore, there is no overlap in thescores of the individuals between the groups.

5. DISCUSSION AND CONCLUSIONS

This section discusses the following: theory of personalization, reliabilityand validity, epistemological premises, design implications, and conclusions.

5.1. The Theory of Personalization of Appearance

The three qualitative studies that have been described have utilized a vari-ety of methods, user groups, and devices. The theory of personalization of ap-pearance that has emerged is summarized in Figure 7. User, system, and con-textual dispositions lead to the personalization of appearance. This hascognitive, social, and emotional effects on the user.

THEORY OF PERSONALIZATION OF APPEARANCE 215

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 26: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

216

ParticipantFrequency

of P (1) Extent of PKnowledge

of P (2)Socioemotional

Context of Use (3)Ownership

of S (4)Dispositional

Index (5)

S. R. 3 Custom and downloaded themes,wallpapers and screensavers.Recorded sounds.

3 5 3 3.7

R. T. 3 Custom and downloaded themes,wallpapers and screensavers. Customand downloaded sounds.

3 5 3 3.7

J. Ma. 2 Screen resolution. 3 4 2 3H. L. 2 Custom and downloaded themes,

wallpapers and screensavers.2 2 3 2.3

W. H. 2 Custom screensaver and wallpaper. 2 3 2 2.3J. M. 1 None. 1 4 1 2I. C. 1 None. 2 2 1 1.7T. S. 1 None. 1 3 1 1.7

Note. S = system or product; P = personalization. (1) 1 = never; 2 = occasionally; 3 = once a month or more. (2) 1 = novice; 2 = intermediate; 3 = ex-pert. (3) 1 = task-based motivation and/or no leisure-based functions in the device; 5 = socioemotional motivation and respective functionality (e.g.,e-mail, games, Internet access) in the device. (4) 1 = participant is not the owner or the main user; 2 = participant feels some ownership; 3 = partici-pant has full ownership. (5) dispositional index (average of three previous columns), see discussion.

Figure 6. Personalization information regarding the participants in Study 3.

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 27: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

One interesting question is whether some kind of positive feedback loopcould arise, such that the effects on the user could reinforce the dispositions topersonalize leading to further personalization, further effects, and so on.Clearly, this is only possible through the user-dependent dispositions. Someof these are relatively unchanging objective states (e.g., someone either is, oris not, the owner of the system or product). As such, the corresponding emo-tional effects feeling in control and feeling of ownership could only cause suchpositive feedback if we were to add some mediating concept such as “percep-tion of ownership.” We have resisted this kind of change to the theory as itconfuses the distinction between dispositions and effects. Other dispositions,however, are more changeable. One could imagine that all of the positive ef-fects on the user could lead to greater frequency of use. Similarly, the SECcould be accommodated through improved aesthetics, and all the social andemotional effects. The possibility of this kind of positive feedback seemshighly plausible, and detailed longitudinal studies of personalization couldyield interesting data on this point.

Another question is to what extent the dispositions are specific to particularuser groups and devices. When comparing the Finnish mobile phone groupswith the two-stage study there were dispositions and effects appearing in one

THEORY OF PERSONALIZATION OF APPEARANCE 217

Dispositions to PersonalizeAppearance

Effects of Personalizationon User

User CognitiveFrequency of use of S Ease of useOwnership of S Improved aestheticsKnowledge of P Recognition of SNew to system

SocialSystem Reflection of personal identity

Ease of P Reflection of group identityCost of PAbsence of technical constraints EmotionalEffectiveness of P items Familiarity with S

S feels personalContext Feeling in control

Socioemotional context of use Feeling of ownershipSeasonal and media influences Release from boredomPeer influence Fun

Positive associationsAttachment to SAccommodating current

emotional state

Note. S = system or product; P = personalization.

Figure 7. The theory of personalization of appearance.

PERSONALIZATION

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 28: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

that did not appear in the other, and vice versa. At the very least this suggeststhat thedeviceandusergroupaffect the relativeweightsof somecategories. Forexample, reflection of group identity and attachment to system or product maybe more salient to Finnish teenagers than UK students. Ease of achieving per-sonalization may be more salient when talking about PCs than mobile phones.These issues can only really be resolved with large formally sampled surveys.Further ideas for future research arise when regarding the general dynamics ofthe theory. For instance, the correlation between the dispositions and actualpersonalization behavior of users should be studied. Our theory predicts that ahigh personalization disposition leads to a high scope of personalization. Ifthese aspects turn out to be related, the theory will receive further support. Aninitial step toward this was already taken in the home PC user study where theassociation between the dispositional index and frequency and extent of per-sonalizationwasexamined.Also, theeffectsofpersonalizationcouldbestudiedby adopting the experimental, Web-based paradigm demonstrated by Blomand Monk (2001). Possible independent variables here could include IT exper-tise, gender differences, age, or usage motivation.

Finally, investigations into the nature of the category improved aestheticsprovide a challenging avenue for future research. This effect is regarded inour theory in a rather unorthodox way, as a cognitive one. Does the casting ofincreased aesthetics as a cognitive effect imply that the appearance of the in-terface places demands on the cognitive resources of the user? Consequently,personalizing the appearance of, say, the desktop, might result in the interfacebecoming “easier on the eye,” that is, less loading on the cognitive resourcesof the user. Conversely, could it be the case that, despite a user’s perception ofthe aesthetics of the system increasing as a result of the personalization pro-cess, the system would actually become more loading than before due to, forinstance, the icons becoming less noticeable? In any case, associating aesthet-ics with cognitive aspects of system usage implies that the aesthetics of the sys-tem competes with other more functional features, such as the amount of in-formation presented on the screen, for the cognitive resources of the user.This notion, in turn, opens up interesting issues for research.

5.2. Relation to Previous Theories

Previous research on personalization of appearance has largely focused onthe area of environmental personalization. Heidmets (1994) provided themost extensive theoretical framework in this area. He concludes that person-alization primarily takes place to control the environment and is accompa-nied by an individualization of the environment by which the participant’sown self is displayed in externalized form. It is interesting to note that theseconclusions correspond to feeling in control and reflection of personal iden-

218 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 29: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

tity in our categories. The fact that these aspects have been found to play arole in environmental personalization and human–computer interactionpoints to the fundamental nature of these behavioral patterns.

Wells (2000) used surveys and interviews to investigate office personaliza-tion and found that personalization and satisfaction with physical work envi-ronment correlated positively. She, however, states that it is difficult to identifya cause and effect relationship between personalization and satisfaction. It maybe that workplaces that allow personalization differ in other positive aspectsfrom those that do not. The causal link between controlling one’s environmentandpositiveemotional responseswasexaminedbyRodin (1986).Elderlynurs-inghomeresidentswereencouraged tomakemorechoicesandhavemorecon-trol of day-to-day events. The group given more control became more alert andactive and reported feeling happier than the low-control group.

Holman (1986) proposed a theoretical framework that highlights the emo-tional ties the consumer has with a product. Five product roles were identifiedand these constituted a dimension of emotional involvement, ranging fromvery low to very high. In these five stages, the status of a product evolves fromthat of a background object to a substitute for human relationships. Interest-ingly for the theoryofpersonalization, the twohighest roles represent situationsinwhich theproduct iseitheranexpressionof selforobjectof theemotion itself.

None of the aforementioned studies consider what we have called thedispositional factors of personalization. This seems to be an important omis-sion if one is trying to make recommendations to designers deciding whetherto include personalization features in systems and products.

Another striking difference between our theory and the others is the largerrange of effects on the user identified. There are various possible explanationsfor this. First there is the method of enquiry. In our use of GTA we have beencareful to keep close to the data. It is possible that some categories could befolded together to provide a simpler theory. Is reflection of personal identityreally a matter of reflecting one’s membership of some subgroup with particu-lar values? Alternatively reflection of group membership may really just bean aspect of personal identity. How real are the differences between the emo-tions listed as effects? Could our participants simply have been struggling todescribe just one or two fundamental emotional experiences? We cannot an-swer these questions; we can only summarize what the participants said. Thetheory in Figure 7 represents the minimum number of categories we wereable to formulate without making these sorts of assumptions.

5.3. Reliability and Validity

Pandit (1996) has made an analogy between a GTA and psychometric testconstruction. In the latter enterprise a test is judged by its reliability and valid-

THEORY OF PERSONALIZATION OF APPEARANCE 219

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 30: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

ity. Reliability refers to whether the results of applying the test are repeatable.Will the same individuals be observed scoring high or low on different admin-istrations of the test or on subsets of the test supposed to measure the samething. Validity refers to empirical arguments about the meaning of the score:Does it measure what it is claimed to measure? In the domain of psychometrictesting there are established criteria for judging reliability and validity. Pandit(1996) suggested these may also be applied to a GTA.

Construct validity refers to how the test, or in this case the theory, was con-structed. It is essentially an argument that the method used was sound so the re-sults should be sound also. The method used here was based closely on Straussand Corbin (1990) and is carefully specified. Internal validity refers to the inter-nalconsistencybetween itemswithina test. In thiscontextonemayaskwhetherthe theory has a consistent structure. The causal model implied by the notion ofdispositions and effects, and the consistent character of the dispositions and ef-fects, can be used to make this argument for internal validity.

External validity would normally refer to a correlation relationship betweenthe test scores and some other independent measure of the same thing. Simi-larities between concepts developed here and in the area of environmentalpersonalization were discussed in Section 5.2. That we were able to applysimilar codes to the discussions of both mobile phones and PCs by samplesfrom such different user populations is also an argument for external validity.Possibly the best demonstration of external validity is the high agreement be-tween the theory and the codes spontaneously generated by the three exter-nal coders in the Finnish mobile phone users study.

Reliability is necessary for validity though the opposite is not the case. Atest may generate repeatable results without meaning what it says it does.However, if it is merely generating random numbers it cannot mean any-thing. Validity then implies reliability; the arguments made, particularly forconstruct and external validity, can also be used to argue for reliability.

It is arguable whether the analogy with psychometric test constructionadds anything to a more conventional justification for the theory. This is that(a) the methods used were systematic and carefully described, (b) the same orsimilar conclusions were drawn on different occasions, and (c) the results ac-cord with a common sense view of what might be expected (after the fact).The most convincing argument for the validity of the theory will be quantita-tive tests of the predictions it makes. Some suggestions about how this mightbe done were made in Section 5.1.

5.4. Epistemological Premises

As discussed in Section 1, one of the assumptions of GTA is that the even-tual theory emerges from the data and is hence not influenced by the precon-

220 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 31: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

ceptions of the researcher (Strauss & Corbin, 1990). This is not to say that pre-vious work should not be consulted when doing GTA. Glaser (1992)suggested that this can be done, but only once the theory has become suffi-ciently developed. Despite referring to external sources at later stages of anal-ysis only, a problem remains. Everyone possesses epistemological premisesthat influence the way they construct the world around them (Burca &McLoughlin, 1996). Thus, any investigator will have some issues to whichthey are sensitive, which will influence the results of the analysis. The “so-cially related motivations” identified in Blom (2000), perhaps guided us tolook for emotional responses and the expression of identity. According toBlumer (1969), this issue is not a problem, however. He refers to the conceptsinfluencing one’s research as sensitizing concepts, which provide the re-searcher with a “general sense of reference and guidance in approaching em-pirical instances.” Thus, sensitizing concepts act as theoretical lenses to helpthe researcher find examples as well as patterns in the meanings representedin data, using theoretical sampling rather than random sampling (Gephart,1999). We were aware of the notion of sensitizing concepts and have takensome trouble to distinguish between these kinds of epistemological premisesand preconceptions, which influence the interpretation of the data in a rathermore negative way. We are confident that we have summarized how our par-ticipants saw the issues in personalization rather than how we saw them, andwe are confident that our previous knowledge has guided the interpretationprocess and has not taken away the groundedness of the theory.

5.5. Design Implications

The theory can inform two related design decisions: (a) What factors willlead users to want to use the personalization features? and (b) Is there a reasonto include features for personalizing the appearance of a particular system orproduct? The former relates most closely to the dispositions and the latter tothe effects. They are related because one important reason to include person-alization features is because users will want them.

Figure 8 presents a checklist, based on the dispositions identified in the the-ory, to determine how likely it is that users will want to personalize a systemor product. If one can answer “yes” to a large proportion of the questions,then the prediction is that there will be a user need for these kinds of features.To illustrate how it could be used, Figure 8 also applies this checklist to threehypothetical examples: teenage users of mobile phones, managers given mo-bile phones for urgent work purposes only, and students using instantmessaging (IM) to keep in touch with family and friends.

The table indicates that, in this hypothetical case, mobile phones, whenaimed at teenage users, fulfill most of the dispositions of the theory of person-

THEORY OF PERSONALIZATION OF APPEARANCE 221

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 32: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

alization. This would point toward the importance of providing the ability topersonalize mobile phones for this particular user group. In addition to identi-fying products with a high potential for personalization, the checklist couldalso be used to identify changes needed to increase the likelihood of userswanting to personalize. Thus, for instance, in the example given earlier, anIM application would have more potential in this respect if the knowledge ofthe ability to accomplish the process was widely available and if it included ef-fective personalization features.

Figure 9 presents a checklist for assessing the need to include features forpersonalizing appearance that is based on the effects in the theory. The de-signer is asked to say if each of these requirements would be one that their tar-get user group would value highly. If some are perceived as being important,

222 BLOM AND MONK

Mobile Phone(Teenager)

Mobile Phone(Manager)

Instant Messaging(Using University PCs)

1. Will the system or product beused frequently by the user?

� �

2. Will the user own the systemor product?

3. Will the user know about theability to personalize and doesshe know how to personalize?

� �

4. Has the user acquired theproduct recently? That is, isthe product new to her?

5. Will it be easy to use thesefeatures?

� �

6. Will the cost ofpersonalization be reasonableto the user?

� �

7. Are the technical constraintsabsent for personalising theproduct?

� �

8. Will the features provide theuser with enough attractiveoptions?

� �

9. Will the context of use be ofsocial and emotional value tothe user?

� �

10. Are there seasonal or mediathemes included in thepersonalization features?

� � �

11. Will there be peer pressureto use these features?

Figure 8. A checklist to determine how likely it is that users will want to personalize asystem or product.

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 33: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

then features for the personalization of appearance with the potential to pro-duce that effect should be considered.

Many products that are currently produced in a single form could in-clude features to facilitate personalization of appearance. The problem forthe designer is to decide when this will provide significant added value tothe user and what form the features should take. The dispositions summa-rized in Figure 7 and codified as checklists in Figures 8 and 9 are a start to-ward answering the question regarding added value. It is much harder togive advice on the latter, what features to design. However, the theory maybe of some help even here by allowing the designer to think clearly aboutthe effects they are trying to achieve. This is only a starting point. Furtherempirical work is needed to test and refine the theory. Also, the issues willbecome clearer as designers and users get more experience of personaliza-tion in future products.

NOTES

Background. The research presented here is associated with Jan Blom’s PhD.Acknowledgments. We wish to thank Jussi Silvonen and Mark Blythe for their ideas

and support on GTA. Gordon Baxter and Paula Lynch have provided valuable assis-tance in various parts of the research. Special thanks to Leena Salo-Gunst and SanttuMerjanaho for their enthusiasm in supporting the Finnish study. We would also like tothank Graham Johnson, Lynne Coventry, and Antonella De Angeli from NCR fortheir contribution to Jan Blom’s research.

Support. Jan Blom’s thesis is supported by NCR Financial Solutions Ltd., UK.Authors’ Present Addresses. Jan Blom, P.O. Box 209, FIN-00045, Finland. E-mail:

[email protected]. Andrew Monk, Department of Psychology, University ofYork, York YO10 5DD, UK. E-mail: [email protected].

THEORY OF PERSONALIZATION OF APPEARANCE 223

1. Ease of use.2. Being able to change the aesthetics of the design to my preferences.3. Being able to recognize my copy of the product or system.4. Being able to reflect my personal identity.5. Being able to reflect my membership of some group.6. Feeling familiar with the system or product.7. The system of product feels personal.8. Feeling in control of the system or product.9. A feeling of ownership of the system or product.10. Being fed up with the appearance of the product or system and wanting to repersonalize it.11. Having fun with the product or system.12. Making me happy.13. Being emotionally attached to the system or product.14. Being able to accommodate my current emotional state.

Figure 9. A checklist of desirable effects for assessing the need to include specific featuresfor personalizing appearance.

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 34: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

HCI Editorial Record. First manuscript received September 10, 2001. Revision re-ceived January 2, 1002. Accepted by Clayton Lewis. Final manuscript received June 4,2002. — Editor

REFERENCES

Asch, S. (1951). Effects of group pressure upon the modification and distortion of judg-ments. In H. Guetzkow (Ed.), Groups, leadership and men (pp. 393–401). Pittsburgh,PA: Carnegie.

Blom, J. (2000). Personalization—a taxonomy. Extended Abstracts of the CHI 2000 Con-ference on Human Factors in Computing Systems (pp. 313–314). New York: ACM.

Blom, J., & Monk, A. (2001). One-to-one e-commerce: Who’s the one? Extended Ab-stracts of the CHI 2001 Conference on Human Factors in Computing Systems (pp.341–342). New York: ACM

Blom, J. (in press). A theory of personalized recommendations. Extended Abstracts of theCHI 2002 Conference on Human Factors in Computing Systems. New York: ACM.

Blumer, H. (1969). What is wrong with social theory? In H. Blumer (Ed.), Symbolicinteractionism: Perspective and methods (pp. 140–152). Englewood Cliffs, NJ: PrenticeHall.

Burca, S., & McLoughlin, D. (1996). The grounded theory alternative in business network re-search [DCUBS research papers 4]. Dublin, Ireland: Dublin City University Busi-ness School.

Carrere, S., & Evans, G. (1994). Life in an isolated and confined environment: A quali-tative study of the role of the designed environment. Environment and Behavior, 26,707–741.

Gephart, R. (1999). Paradigms and research methods: Academy of Management Online,Research Methods Forum (4). Retrieved May 1, 2001, fromhttp://www.aom.pace.edu/rmd/1999_RMD_Forum_Paradigms_and_Re-search_Methods.htm

Glaser, B. (1992). Basics of grounded theory analysis. Mill Valley, CA: Sociology.Goffman, E. (1959). The presentation of self in every day life. Middlesex, England:

Harmondsworth.Guttman, R., Moukas, A., & Maes, P. (1998). Agent-mediated e-commerce: A survey.

Knowledge Engineering Review, 13(2), 147–159.Heidmets, M. (1994). The phenomenon of personalization of the environment: A the-

oretical analysis. Journal of Russian & East European Psychology, 32, 41–85.Holahan, C., & Saegert, S. (1973). Behavioral and attitudinal effects of large-scale vari-

ation in the physical environment of psychiatric wards. Journal of Abnormal Psychol-ogy, 82, 454–462.

Holman, R. (1986). Advertising and emotionality. In R. Peterson, W. Hoyer, & W. Wil-son (Eds.), The role of affect in consumer behavior (pp. 119–140). Lexington MA:Lexington Books.

Kasesniemi, E., & Rautiainen, P. (2002). Mobile culture of children and teenagers inFinland. In M. Aakhus & J. Katz (Eds.), Perpetual contact: Mobile communication, pri-vate talk, and public performance (pp. 170–192). Cambridge, England: CambridgeUniversity Press.

224 BLOM AND MONK

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 35: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

Kobsa, A., Koenemann, A., & Pohl, W. (1999). Personalized hypermedia presentation tech-niques for improving online customer relationships (66). Sankt Augustin, Germany:GMD—Forschungszentrun Informationstechnik GmbH.

Pandit, N. (1996). The creation of theory: A recent application of the grounded theory method.Retrieved May 1, 2001, from http://www.nova.edu/ssss/QR/QR2-4/pandit.html

Peppers, D., & Rogers, M. (1997). Enterprise one to one: tools for competing in the interactiveage. New York: Currency Doubleday.

Rodin, J. (1986). Aging and health: effects of the sense of control. Science, 233,1271–1276.

Schachter, S., Ellertson, N., McBride, D., & Gregory, D. (1951). An experimentalstudy of cohesiveness and productivity. Human Relations, 7, 403–439.

Scheiberg, S. (1990). Emotions on display: the personal decoration of workspace.American Behavioral Scientist, 33, 330–338.

Strauss, A., & Corbin, J. (1990). Basics of qualitative research. London: Sage.Vinsel, A., Brown, B., Altman, I., & Foss, C. (1980). Privacy regulation, territorial dis-

plays, and effectiveness of individual functioning. Journal of Personality and SocialPsychology, 39, 1104–1115.

Wells, M. (2000). Office clutter of meaningful personal displays: The role of office per-sonalization in employee and organizational well-being. Journal of EnvironmentalPsychology, 20, 239–255.

APPENDIX A: STUDY 1 FOCUS GROUP PROTOCOL

A. Initial Stage

• I know that at least we’re all university students. Perhaps it’s a good ideato start by introducing ourselves and maybe what course we’re each do-ing.

B. Introductory Questions

• What do you use the Web for normally?• How do these examples, Excite and Moviecritic, compare to your expe-

rience?

C. Transition Questions

• Tell me about the process of personalizing Excite. How did it make youfeel?

• What where your feelings toward Excite the first time you visited it afteryou had personalized it?

• Would you personalize it further?• What about Moviecritic? What was it like to fill in the movie ratings?

THEORY OF PERSONALIZATION OF APPEARANCE 225

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 36: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

• And how did it feel to actually use the site for the first time to receive rec-ommendations?

• Would you use this in the future? Why?• Provide an explanation of the definitions of personalization (both user

and system initiated).• Which kind of personalization do you prefer? Why?

D. Key Questions

• I’ve got a pen and paper here for you. Please write down: What wouldyou like to personalize? Why?

• Write down (a) What area would you like to receive recommendationson? (b) Why you would wish to receive personalized recommendations?

• How does the fact that the site uses other people’s recommendationsmake you feel?

E. Ending Questions

• If you had a chance to influence the designer of a site, how would youadvice them to use these personalization features?

• I’m aiming to find out your perceptions on personalization. Have Imissed anything important?

APPENDIX B: THEORY OF PERSONALIZATIONCATEGORIES

226 BLOM AND MONK

Study 1

Category InstancesGroups(Total 7) Sample Extract

Frequency of useof system

22 6 I would personalize my e-mail because you seeit so often.

Ownership ofsystem

18 6 If something is my own I’d like to change it.

Knowledge ofpersonalization

13 3 I don’t think I would have done it had I notbeen asked to. I don’t think I would haveeven noticed that you could have done it.

Ease ofpersonalization

4 3 [What would you like to personalize?] Justthings on my mobile phone and my laptop. Ilike doing it because they are all easy tochange.

(continued)

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 37: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

THEORY OF PERSONALIZATION OF APPEARANCE 227

Study 1 (Continued)

Category InstancesGroups(Total 7) Sample Extract

Effectiveness ofpersonalizationfeatures

7 3 I like the idea of personalization but only if it’smore specific to me. I’m not really interestedin most of the stuff. It’s a good idea but it’snot specific enough for what I wan to see.

Socioemotionalcontext of use

7 3 It’s like on word when you sit down and writean essay or something and you don’t wantthose colorful flashy things distracting you.But with e-mail it’s not just work. You’rewriting to friends and things.

Ease of use 12 4 I always put bright colors in … it’s the ease ofuse, and speed up.

Improvedaesthetics

18 5 It’s probably the same as if you had an officeand you put a picture on the wall. Just tomake it look a bit nicer.

Recognition ofsystem

7 3 In my house, there’s six of us who have thesame phone…And if you got a different ringthen that helps.

Familiarity withsystem

9 4 When you’ve personalized something you justinstinctively know what this thing is.

System feelspersonal

14 5 I liked that [personalization of Excite], your ownlittle greeting that made it a bit morepersonal.

Feeling in control 8 4 [Personalization] Makes you feel more incontrol so you actually know what you’redoing.

Feeling ofownership

18 6 It makes it yours [personalization]. It’s like oncampus, in your room, you stick up postersup don’t you.

Release fromboredom

5 4 I just like blue and I was bored so I changed it.

Fun 9 3 I have a U2 desktop theme on my computer. Itplays little tunes to me. It’s quite amusingreally.

Positiveassociations

15 7 I like to scan photos of places I’ve been to if I’vehad really good holidays. Cause I just look atit and … it brings back good memories and ithelps me feel generally positive.

Reflection ofpersonal identity

8 5 Everyone wants to see an identity, their ownidentity, have this stuff identified. This is minecause it’s got that on it, it looks like that.

Reflection of groupidentity

7 3 My best friend sent me a “Barbie girl” themetune cause I’m dizzy. So I had that as myringing tone for quite a while. Quite a goodtalking point, “why have you got that?”

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14

Page 38: Theory of Personalization of Appearance: Why Users Personalize Their PCs and Mobile Phones

228 BLOM AND MONK

Study 2

Category InstancesGroups(Total 3) Sample Extract

Accommodatingcurrent emotionalstate

10 3 The ringing tones reflect my mood.

Seasonal and mediainfluences

5 3 Quite a few people under Christmas timeorder “Jingle Bell” stuff to their phones.

Peer influence 27 3 [What makes you personalize?] Someone hasput in something interesting [operator logo],so I’m like: “OK, I’ll go and get one too.”

Cost ofpersonalization

6 3 It’s easy to buy mobile phone accessories, asthey are not so expensive.

Absence of technicalconstraints

9 3 It annoys me that I cannot send logos to myphone.

Attachment tosystem

17 3 It’s like your [personalized phone] own little,not an animal, but your “little one.”

Study 3

Category InstancesPersons(Total 8) Sample Extract

New to system 5 4 I did use to [personalize], when I first got thecomputer, I used to fiddle about with theappearance all the time. Just because it was anew toy. Get acquainted a bit.

Dow

nloa

ded

by [

Tow

son

Uni

vers

ity]

at 1

8:29

13

Nov

embe

r 20

14