16
Participant observation in logistics research Experiences from an RFID implementation study Henrik Pa ˚lsson Division of Packaging Logistics, Department of Design Sciences, Lund University, Lund, Sweden Abstract Purpose – The purpose of this paper is to describe the advantages, challenges and uncertainties of collecting and analyzing data using participant observation in logistics research. Design/methodology/approach – Experiences from a participant observation study of an interorganizational radio frequency identification (RFID) implementation in an international environment are presented and reflected on. The RFID implementation included complex interactions between three leading companies. Findings – The results appear to support an increased use of participant observation in qualitative logistics research, particularly when investigating interorganizational aspects. The analysis highlights values, general limitations and challenges of using participant observation in logistics. The paper illustrates that using participant observation results in significant and detailed findings, which would be difficult to achieve with other methods. Suggestions on how to take advantage of the method’s benefits and overcome methodological challenges are provided. Research limitations/implications – Future research may address experiences from other studies regarding how to analyze and report data from a participant observation study. It may also clarify the role the method is given in case studies and extend the analysis of epistemological aspects conducted in this paper. Practical implications – This paper may inspire logistics researchers to consider participant observation, either as sole method or as part of a multi-methodical case study, in order to make use of its benefits and thus broaden the dimensions of logistics research. Originality/value – A broad literature review indicates that participant observation studies are rather uncommon in logistics research. This paper thus highlights the potential of using this method in logistics research, particularly when investigating the overlooked, but essential, interorganizational aspects of logistics and SCM. Keywords Quality, Research, Participative management, Logistics data processing Paper type Research paper Introduction When writing a paper considering a participant observation study on a radio frequency identification (RFID) implementation project, the author was going through leading journals within the field of logistics to gain inspiration from other participant observation studies. Surprisingly, only a limited number of logistics studies using participant observation were found. In addition, examination by Mentzer and Kahn (1995) and by Sachan and Datta (2005) confirmed the author’s suspicion that participant observation studies within logistics are rare. In the comprehensive analysis by Mentzer and Kahn, participant observation is not even a category. Perhaps, it is The current issue and full text archive of this journal is available at www.emeraldinsight.com/0960-0035.htm IJPDLM 37,2 148 International Journal of Physical Distribution & Logistics Management Vol. 37 No. 2, 2007 pp. 148-163 q Emerald Group Publishing Limited 0960-0035 DOI 10.1108/09600030710734857

Participant observation

Embed Size (px)

DESCRIPTION

 

Citation preview

Participant observation inlogistics research

Experiences from an RFIDimplementation study

Henrik PalssonDivision of Packaging Logistics, Department of Design Sciences,

Lund University, Lund, Sweden

Abstract

Purpose – The purpose of this paper is to describe the advantages, challenges and uncertainties ofcollecting and analyzing data using participant observation in logistics research.

Design/methodology/approach – Experiences from a participant observation study of aninterorganizational radio frequency identification (RFID) implementation in an internationalenvironment are presented and reflected on. The RFID implementation included complexinteractions between three leading companies.

Findings – The results appear to support an increased use of participant observation in qualitativelogistics research, particularly when investigating interorganizational aspects. The analysishighlights values, general limitations and challenges of using participant observation in logistics.The paper illustrates that using participant observation results in significant and detailed findings,which would be difficult to achieve with other methods. Suggestions on how to take advantage of themethod’s benefits and overcome methodological challenges are provided.

Research limitations/implications – Future research may address experiences from other studiesregarding how to analyze and report data from a participant observation study. It may also clarify therole the method is given in case studies and extend the analysis of epistemological aspects conductedin this paper.

Practical implications – This paper may inspire logistics researchers to consider participantobservation, either as sole method or as part of a multi-methodical case study, in order to make use ofits benefits and thus broaden the dimensions of logistics research.

Originality/value – A broad literature review indicates that participant observation studies arerather uncommon in logistics research. This paper thus highlights the potential of using this method inlogistics research, particularly when investigating the overlooked, but essential, interorganizationalaspects of logistics and SCM.

Keywords Quality, Research, Participative management, Logistics data processing

Paper type Research paper

IntroductionWhen writing a paper considering a participant observation study on a radiofrequency identification (RFID) implementation project, the author was going throughleading journals within the field of logistics to gain inspiration from other participantobservation studies. Surprisingly, only a limited number of logistics studies usingparticipant observation were found. In addition, examination by Mentzer and Kahn(1995) and by Sachan and Datta (2005) confirmed the author’s suspicion thatparticipant observation studies within logistics are rare. In the comprehensive analysisby Mentzer and Kahn, participant observation is not even a category. Perhaps, it is

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/0960-0035.htm

IJPDLM37,2

148

International Journal of PhysicalDistribution & Logistics ManagementVol. 37 No. 2, 2007pp. 148-163q Emerald Group Publishing Limited0960-0035DOI 10.1108/09600030710734857

included as a part in some of the case studies reviewed, but only about 3 percent of thearticles published in the Journal of Business Logistics 1978-1995 were case studies.Sachan and Datta (2005, p. 666) pointed out that existing reviews show that directobservation methods are rather unpopular and that “researchers are mainly usingpeoples (sic) perception (survey and interview) or artificial methods (simulation andmathematical modeling) for research in the discipline.”

To further examine this issue, the author conducted a general search in our internallibrary navigator, which contains most of the logistics journals relevant to this areaand a considerable number of conference proceedings. A search for participantobservation or ethnographic studies in logistics or supply chain management resultedin only a handful of matches. This apparent lack of participant observation research isconfirmed in a review of case study research (Seuring, 2006). Only five of 68 casestudies employed participant observation in any way. Thus, it would not beexaggerating to say that participant observation is a rarely used method in logisticsresearch.

However, Sachan and Datta, as well as Naslund (2002), clearly state that logisticsneeds more qualitative research. Sachan and Datta (2005, p. 669) point out the needfor movement toward direct observation via case, action, and field studies, as“the methods are accessible, their legitimacy is proven, and the need is great.” Naslundemphasizes the need for more ethnographic studies and action research.

This paper is thus designed to strike a blow for an increased use of participantobservation in logistics research. Therefore, a study on RFID implementation usingthis method is offered to illustrate the potential of the method in logistics.

The purpose of this paper is to describe the advantages, challenges, anduncertainties of collecting and analyzing data using participant observation in logisticsresearch. Experiences from the interorganizational RFID implementation study in aninternational environment are presented and reflected on. The paper also brieflydiscusses whether existing epistemological settings in logistics support or areobstacles for participant observation.

The outline of the paper is the following. It begins with a review of participantobservation and positioning of the method compared to other qualitative methods.Then the interorganizational RFID implementation study is summarized, with a focuson methodological aspects. Thereafter, the study is reflected on. Subsequently,participant observation in logistics is discussed from a more general perspective, butwith connections to the study carried out. Finally, future research is suggested.

Participant observationCharacteristics of participant observationDifferent authors seem to include various aspects when they address participantobservation. Some literature (Jackson, 1983; Park, 1999) regards it as a rather broadresearch strategy, including observations, interviews, and sampling from documents.These authors almost place it on a par with ethnography, while other literature(Merriam, 1994; Yin, 2003) has a narrower view of it. Yin considers it as more of a datacollection technique which can be used within case studies. Merriam, on the other hand,regards it as a method and he claims it to be one of the most important methods in casestudies. Bryman (2002) also considers it to be a method, while he addressesethnography as a wider term. In logistics research, Yin’s approach of regarding

Participantobservation in

logistics research

149

participant observation as a data collection technique, often as a part of a case study,appears to be common. However, participant observation as the main research methodin logistics is also available (Ellram, 1996), which is the view adopted in this paper.

Participant observation is defined by Bryman (2002). He states that aparticipant observer is engaged in a group for a considerable period of time. Thebehavior of the group is explored by observing conversations within the groupand with the researcher. Bryman also points out that it is common that participantobservation also incorporates supplementary interviews and written material.

To distinguish participant observation from other qualitative methods, it iscompared to five related methods. Ethnography may refer to either a method or aphilosophical paradigm (Naslund, 2002). This method is close to participantobservation, but it is often more culturally focused (Bryman, 2002).

Participatory action research has similarities to participant observation, as a groupis investigated over a period of time. It does, however, significantly differ from thismethod on the point that “some of the people in the organization or community understudy participate actively with the professional researcher throughout the researchprocess” (Foote Whyte, 1991, p. 20).

Case studies may be defined as an investigation of “a contemporary phenomenonwithin its real-life context” (Yin, 2003, p. 13) or as the study of the complexity and natureof a specific case (Stake in Bryman, 2002). A case may, for instance, be a program, anevent, a process, an institution, or a social group (Creswell, 1994). According to Raginand Becker (Ragin, 1992, p. 225) it is “a way station in the process of producing empiricalsocial evidence.” In a case study this evidence may be produced from data collected byusing various methods, both qualitative and quantitative ones (Ellram, 1996); hence,participant observationmay be one of them. Accordingly, case study researchmay haveoverlapping parts with participant observation, for instance, collecting empiricalevidence from a contemporary phenomenon, interviewing, and observation. Participantobservation as viewed in this paper, however, is more distinct, focusing on spendingtime in a studied group, and if interviews are used, they are complementary or lessformal. This means that interviews are mainly conducted for unobservable parts of thestudied phenomenon, while a case study may use them as the main method.

Interview studies are a stand-alone method. Compared to participant observation,one basic and essential difference is that participant observation mainly results infirst-hand data from a contemporary phenomenon, while interviews result in datainterpreted by the respondent from a historical event. A thorough comparisonis provided by Bryman (2002). Participant observation offers better opportunities toreveal tacit knowledge and unexpected behavior and is thus able to adapt to changedsituations. It also has advantages in being able to uncover hidden activities andhighlight contextual sensitivity. Interview studies, on the other hand, are better atexplaining certain types of emotions and underlying causes for some actions, forexample, why a person has become a vegetarian. Participant observations are suitablefor in-depth longitudinal studies of a limited period of time, but for practical reasonsinterview studies are more suitable for long, longitudinal investigations and historicalevents. Another practical advantage of interviews is that they often requireconsiderably less time. Interview studies also offer a wider scope of the study as theymay include people from various parts of an organization, while participantobservation is often limited to a small part of an organization.

IJPDLM37,2

150

Focus groups represent a certain type of interview where several respondentsdiscuss questions asked by the researcher. Consequently, the comparisons betweeninterview studies and participant observation are, to a large extent, also valid for focusgroups. However, there are some characteristics which put focus groups in anotherposition than interviews compared to participant observation. In accordance withparticipant observation, focus groups result in large amounts of data which may bechallenging to analyze. Compared to interviews, focus groups offer slightly lesscontrol, but not to the extent that is typical in participant observation. Whileparticipant observation reveals the first-hand data of a phenomenon, Bryman (2002)states that a focus group may come to consensus when an issue is being discussed.A risk of this course of action is that people in a group stop thinking critically. Anotherrisk put forward by Bryman is that group effects may limit the results of a focus group,as some participants may be loud and some very quiet. These risks are considered tobe less apparent in participant observation because the researcher studies the group inits natural context.

Four phasesFour phases (preparation, data collection, analysis, and writing) which could beidentified in ethnographic studies are also valid to some extent in participantobservation studies. They are referred to below.

The preparation phase includes, similar to other research methods, research design.An essential element is gaining access to the objects to be investigated. This often includesseeking the permission of gatekeepers and the support of sponsors. However, access is notonly a question of physical presence or absence (Hammersley and Atkinson, 1995, p. 55).Private boundaries may, for example, be difficult to break through, as they “may bepoliced by gatekeepers” (Hammersley and Atkinson, 1995, p. 63).

Further, in the field relations it is essential to gain the trust of the group beingresearched. Thus, Hammersley and Atkinson (1995) observe that it may be necessaryto dress similarly to the group and to clearly explain the intentions of the study. Theyalso recommend a researcher not to become too involved in a group if it preventshim/her from gaining access to other groups.

After preparation, the data collection phase may begin. Various authors offerpractical advice on how to ensure accurate and trustworthy information. Glaser (1996),for instance, claims that it is essential to:

. Schedule strategically – be flexible, take advantage of “snowballing”(one respondent may suggest and facilitate access to other respondents).

. Adjust the interview request to the situation – for example, taking advantage oftime in between events to conduct short interviews.

. Obtain multiple perspectives.

. Establish trust and gain access.

. Reveal your purpose, suppress your opinion – be clear, but do not reveal youropinions on different issues.

. Do not overlook detail in the rush of activity – it is impossible to knowbeforehand which details are critical in the investigation.

. Be a pack rat – it is impossible to know beforehand what is important.

Participantobservation in

logistics research

151

Jackson (1983), on the other hand, identified six challenges:

(1) Problems of data handling – for example, sorting very comprehensiveinformation.

(2) Significant time gap between the occurrence of an event and its being recordedas data.

(3) Lengthy delay between research and writing.

(4) Problems of systematic analysis and convincing presentation of infiniteamounts of data.

(5) Too heavy reliance on informants involves the dangers of an elite bias.

(6) Ethical and moral questions.

Atkinson et al. (2001) put emphasis on field notes. They state that field notes may beeither jotted down or merely mental. The jotted notes may work as reminders forcertain situations or may be the initial steps of the writing process. Mental notes are tobe conducted while the researcher is observing in detail, with the aim of writingor participating in ongoing events in order to gain experience of certain situations orprocesses.

The phase of analysis and reflection in participant observation studies is typicallyongoing throughout the investigation. Hammersley and Atkinson (1995, p. 210) offersome advice on how to conduct the analysis in ethnography, which is useful inparticipant observation too; “The first step in the process of analysis is, of course,a careful reading of the corpus of data, in order to become thoroughly familiar with it.”They point out that the aim of the first step is to identify patterns, seek relationshipsacross the whole data or see whether anything stands out. They further explain thatconcepts may also be observed. The initial analysis is expected to generate fairly trivialconcepts which, further on in the analysis, become more abstract. A more in-depthanalysis strategy, suggested by Hammersley and Atkinson (1995, p. 216), is constantcomparison (Glaser and Strauss, 1967). Another strategy they propose is typologydevelopment, where “an initial set of categories differentiating a particular range ofphenomena can be developed into a systematic typology.” Irrespective of the strategychosen, the interpretations in the study depend on several aspects:

. Social context. Since, the participant observer interacts in a social context, theaudience “to which the actions or accounts being used as data were directed”(p. 220) should be considered, as this may affect what is being said. It should betaken into account that the people being studied may have something to gain bynot telling the truth. The audience may be the researcher, for instance, ininterviews and some observation situations.

. Time. What is said and done depends on time, i.e. what has already occurred isrelevant to decisions and actions.

. Personnel. Actions and perspectives of people depend on their identity and theirsocial relationships. For example, the position a person holds in a companydetermines which information is accessible to them.

. Respondent validation. The respondent may have more knowledge than theresearcher, but the respondent’s knowledge may be false or there may be amotive behind an incorrect description or a misinterpretation. If the investigation

IJPDLM37,2

152

may be interpreted as critical or negative, respondent validation may also beproblematic.

. Triangulation. Checking findings with different sources.

Another way of analyzing data in a participant observation study is to use codingtechniques from grounded theory. Strauss and Corbin (1998) emphasize the need tocarefully scrutinize data. Accordingly, they describe how to break down data withextensive coding procedures. They also promote the writing of memos and notes togain analytical distance from data.

An essential part of any analysis is how data are interpreted. Alvesson andSkoldberg (1994, p. 12) put the emphasis on reflection, which they define as“interpretation of an interpretation.” This means that reflection aims to criticallyevaluate interpretations of empirical material conducted by a researcher. Thus, theconnection between interpreted data, for example, models, and reality in data may bestrengthened.

Finally, the phase of writing may be divided into two parts; writing field notesand writing the finished texts. This phase is more extensive in ethnographic studiesand thus only briefly mentioned here. Atkinson et al. (2001) point out that writing fieldnotes often is a first analytic step. It may facilitate new insights into the research areaand give rise to new ideas. In the final text in a participant observation study, the fieldnotes are reported together with other empirical data, for example, interviews.

Participant observation in the RFID implementation studyA brief overview of RFID technology in logisticsBefore highlighting methodological issues of the participant observation studyconducted, RFID technology in logistics, its potential advantages and challenges arebriefly presented. Without visual contact RFID technology captures data from anobject. In a typical RFID system, a reader transmits and registers radio waves whichare modified by a tag (antenna) applied to an object.

The potential of using RFID technology in logistics has attracted a great deal ofattention recently (Sheffi, 2004). Current literature has, to a great extent, focused on thepotential opportunities this technology offers in logistics. Potential advantagesreported are reduced labor hours (Jones et al., 2004; McFarlane and Sheffi, 2003), lessspoilage (Karkkainen, 2003), and a reduction in shrinkage in the supply chain (Joneset al., 2004; Rutner et al., 2004). Other potential advantages identified are increasedprocess efficiency (Angeles, 2005; Rutner et al., 2004), improved sales due to reducedout-of-stock situations (Jones et al., 2004; Prater et al., 2005), improved track and traceopportunities (Angeles, 2005; Jones et al., 2004), improved accuracy of product control(Angeles, 2005; Rutner et al., 2004), and improved inventory management in vendormanaged inventory settings (Smaros and Holmstrom, 2000).

Two generic types of logistics systems using RFID exist. First, RFID may beimplemented in a closed loop, whichmeans that the same RFID tag is used over and overagain in a closed loop. Since, the tag is used many times, its purchase price is relativelyunimportant. Consequently, rather expensive high-performing tagsmay be used in suchloops. Second, RFID may be implemented in an open system, for example, in a supplychain. In this context, several challenges arise. There are technological challenges(McFarlane andSheffi, 2003); the tags are disposable andmust be inexpensive. Thus, it is

Participantobservation in

logistics research

153

challenging to produce tags which are inexpensive but still reliable and durable.In addition, implementation in a supply chain should also consider the technologicalrequirements of several actors; the RFID system chosen should work for all actorsinvolved. Furthermore, this interorganizational character of RFID implementation in asupply chain results in other challenges. Interorganizational barriers are associatedwithcost/benefit sharing, collaboration, information sharing, and technology transfer issues(Palsson, 2006).

Nonetheless, RFID initiatives are currently being mandated by large retailers, forexample, Wal-Mart, Tesco, and Metro. Therefore, their top suppliers are more or lessrequired to implement RFID technology to be used in the supply chain.

Overall description of the RFID implementation studyTo substantiate the methodological analysis, a brief description of the RFID studyconducted is offered. The description reveals social context and personnel aspectswhich influenced the interpretations of the study. The investigation examined aninterorganizational implementation project of RFID technology in disposablesecondary packaging; the project was mandated by a large retailer. The studyfocused on interorganizational supply chain relationships and emphasized theimplications of an RFID implementation project. It highlighted the interactions of aproject group consisting of representatives from a packaging supplier, a foodmanufacturer, and a technology provider.

The unit of analysis was thus the packaging company with the focus on itsinteractions with the participants in a working group aiming at implementing RFIDtags in disposable secondary packaging. This case was chosen because it represents aleading packaging supplier and a leading food manufacturer with high levels oftechnological development. Both companies aim to be at the forefront of newtechnological concepts and tools in their industry.

The nature of the companies involved varied somewhat. Those investigated at thepackaging company were head office representatives. This company is ranked as oneof the top companies in Europe regarding sales and market shares. The packagingcompany’s customer who was involved in the study was a very large foodmanufacturer. Representatives were located at the head office. The representatives ofthe third company in the study, the technology provider, were part of the company’sEuropean subsidiary. This company is much smaller than the other companies.

The relationship in the working group was interorganizational, which addedcomplexity to the study. As the relationship was interorganizational, power wasidentified as impacting the collaboration substantially, both in terms of coercive andreward power (French and Raven, 1958). It was possible to initiate the project due tothe power wielded by both the retailer and the food manufacturer. The main reason forthe packaging supplier to enter the project was fear of reprisals, but while participatingin the project the company also realized the potential of RFID technology for addedbusiness opportunities. The complexity of the study was also affected by four essentiallinks between the collaborating parties. The links expressed mutuality (Dubois andGadde, 2000). There were activity links between the companies both operationallyand strategically, resource ties in terms of shared resources in testing and conductingthe project, continuous interaction in a working group and, finally, economic linksregarding implementation costs. Another complexity enhancer was the fact that the

IJPDLM37,2

154

collaborating companies had varying goals for participating. Finally, complexity wasalso added by a need to share costs in the interorganizational collaboration.

However, besides being interorganizational, the study dealt with RFID technologywhich was an unproven technology interorganizationally. This made the complexity inthe study even more evident.

The investigation identified a number of characteristics of the relationship, andessential issues which needed to be considered in implementing RFID technology in asupply chain. It revealed a need for knowledge of both packaging and the RFIDtechnology. Moreover, it demonstrated the impact of power on the relationship in theinterorganizational RFID implementation project as well as reasons for uncertainty inthe implementation project. It was suggested that uncertainty could be reduced with aproactive approach, timely reporting, assigning clear roles, and early agreement of costsharing. The study also identified a lack of and a need for both a common goal and acost-sharing strategy. An agreed common goal appeared to improve project efficiency.

PreparationThe preparation phase included two main elements; designing the study and gainingaccess. Participant observation was selected as the main research method for severalreasons. It made it possible to study a contemporary phenomenon and thus gainotherwise inaccessible information (Yin, 2003) as well as to receive firsthand, detailedinformation about the phenomenon. It also facilitated a holistic interpretation of thesituation. The research design further included elements such as definitions of periodof time, purpose and scope of the study, and key people and organizations in the study.Gaining access needed particular attention. For initial skepticism to be overcome fourcomponents were identified as helpful. First, the author had initial contact with one ofthe organizations to be investigated. Thus, it was decided to first come to an agreementwith this organization to use its staff as sponsors later on. This proved to be importantas the sponsors had a key role in convincing the most skeptical project group members.Second, a thorough description of the background of the author led to both a betterworking relationship and helped convince the investigated group that the author couldcontribute to the implementation process. Third, once the purpose of the study wasclearly revealed, it was easier for it to be accepted. Finally, making clear that the authorwould contribute to the progress of the implementation also facilitated the projectgroup’s acceptance of a researcher as a member.

The process of data collection and data analysisParticipant observation was the main element of a single case study on implementationof RFID technology in a German retail supply chain (Figure 1). The participantobservation was mainly conducted in a working group and in an experiment.

Figure 1.Methods used in the

interorganizational RFIDimplementation study

Case study

Interviews

Participantobservation

Working groupExperimentStudy visit

Participantobservation in

logistics research

155

The phase of data collection began when an interorganizational project regardingimplementation of RFID technology in the supply chain (Figure 2) was started.The main part of the project was an active working group, consisting of six permanentmembers, including the author, originating from the packaging supplier, the foodmanufacturer, the technology provider, and academia. Temporary members wererepresented as knowledge resources when needed. The working group had aninternational character as the permanent members represented three nationalities andthe temporary members another two. The role of the author was mainly to documentand observe the progress of the project and secondarily to participate in discussionsand some project tasks.

The overall goal of the implementation project was somewhat unclear. Given thecomplexity of the study, this is unsurprising. However, the goal was interpreted bythe working group as making an RFID system work on secondary packaging, i.e. allsecondary packages should be labeled individually with RFID tags and all tags shouldbe read simultaneously on several places in the supply chain. Thus, in order to furtherevaluate the new technology, the working group identified two vaguely defined tasks.First, a working, robust RFID system should be chosen, including both technologicalchoices such as RFID tags, positioning of tags on packages, readers, effects, etc. andlogistical effects such as supply chain design, efficiency improvement, reading locations,etc. Second, a financial evaluation of introducing RFID technology was needed.

The working group existed for approximately six months. After a kick-off inSeptember 2005 the group operated on a regular basis until February 2006. The closingof the working group was rather turbulent and abrupt as one of the project leaders leftthe packaging company due to downsizing. This led to a hiatus in the project.

One type of knowledge resource used in the project was study visits at companieswith knowledge of RFID technology. This facilitated the working group’sunderstanding of the technological requirements and limitations of RFIDtechnology. It also served as inspiration in the process of considering futureopportunities of RFID technology.

Another knowledge resource utilized was that of experiments conducted at thetechnology provider’s. Together with the technology manager, the author carriedout technology tests of RFID technology. The goal was to reach a high read-rate forRFID tags in packaging. Each transportation package on a pallet had a unique RFIDtag, and as the pallet was pulled between two readers all tags should be read. To obtaina 100 percent accurate read rate we had to find a good combination of type (brand) andposition of tag, as well as type, frequency, and effect of readers.

Data were thus collected using participant observation in three general types ofenvironments. In these environments the level of participation/observation varied(Figure 3). During the working group meetings the focus of the researcher was

Figure 2.Actors involved in theRFID project

Packaging supplier

Food manufacturer

Technology provider

Knowledge resources

Retailer

IJPDLM37,2

156

primarily to observe, and secondarily to participate. The observation regardeddocumentation and interpretation of the project progress reported and discussionscarried out in the meetings. To fully make use of these events the author complementedthe observations with informal interviews or discussions with the participants.The second type of environment for data collection was that of experiments. Here, thesituation required the author to take a fairly active role due to lack of resources in thefield of RFID knowledge. The RFID testing was carried out by the technology managerat the technology provider’s and the author. The final environment for data collectionwas that of study visits. Since, the overall aim of the whole working group was toobserve, the observer role of the author was quite natural. To gather even more data,semistructured interviews were conducted. Thus, it was possible to obtain reflectionsfrom both working group members and external interested parties.

The phase of analysis was to some extent ongoing throughout the investigation.The aspects which would influence data interpretation were applied in various degrees.Most data were written in documents and were continuously interpreted and analyzed.The information was analyzed through coding, inspired by grounded theory (Glaserand Strauss, 1967; Strauss and Corbin, 1998). The information was structured andanalyzed both chronologically, to take the time aspect into account, and according tocontent. To emphasize the need for variation and width in the interpretations,interpretative reflection (Alvesson and Skoldberg, 1994) was applied in the analysis.Conducting a reflexive analysis required a careful, detailed use of, and reflection on, theempirical material involved (Thomsson, 2002). Thus, the interpretations werecompared to theories and the author’s prior understanding of the subject. It was alsopossible to take into account the aspects of social context and personnel. Respondentvalidation was only considered to a limited extent, because the study was slightlynegative and hence this aspect is judged to be problematic. Instead, another reflectionwas applied using triangulation, i.e. one part of the empirical material wasstrengthened in other parts of the empirical material.

Reflections on the study – pragmatic issuesReflections on pragmatic issues from the RFID implementation study are presentedhere to illustrate the real effects of the methodological choice. This illustrationfacilitates a concrete evaluation of the advantages, challenges, and uncertainties ofparticipant observation in logistics. Thus, the methodological aspects of participantobservation in logistics investigated are firmly tied to empirical data on a pragmaticlevel. The reflections are mainly directed at the phases where the current studyidentified the main challenges, which were in data collection and analysis. Thesephases are not explicitly pointed out in this reflection as they interacted throughoutmost of the study.

Figure 3.Role of the researcher in

different parts of theresearch

ParticipationObservation

Study visitWorking

groupmeeting

Experiment

Participantobservation in

logistics research

157

Methodological pros in the RFID implementation studyA critical issue in participant observation is gaining access. Access to the workinggroup was difficult to obtain, but once gained new opportunities arose. It providedopportunities to gather information from busy people through informal interviews.The relationship between the working group members also gave insight into sensitiveoccurrences behind the scenes. For example, it was found that internally at thepackaging supplier’s, nobody was really interested in the project. Once this had beenrealized, its effect on the project could be observed. A lack of goals, for instance, wasrecognized and investigated further. Another occurrence behind the scenes whichwould have been tricky to identify without observation was how the cost-sharingstrategy was carried out by one of the companies in the project. By consistentlyignoring the issue of cost sharing, one of the participating companies postponed thisissue. On occasions, when the matter was brought up, the company hushed it up.

By gaining access to a group, participant observation can give insight intootherwise inaccessible firsthand information. In the RFID study, it was found that theparticipant organizations in the working group had different goals. As a result, thisdiversity and its effect were studied and analyzed. Detailed firsthand informationregarding a minor conflict which arose regarding cost sharing was also accessed.By participating and observing it was possible to directly discover several areas ofuncertainty in the project. Part of the uncertainty, for example, unstructuredexperiments, an ad hoc approach, and vague responsibilities, would not have beenrecognized through using another method.

Comprehensive amounts of data may improve the quality of a study. Here, anin-depth understanding of the implementation process was offered, as the methodfacilitated detailed data collection. Interesting information was, for example, gained bythe author studying mail correspondence in the working group, discussions, andworking group meetings. This information facilitated an analysis of the relationship inthe working group identifying, for example, what was lacking for an effectiverelationship to be assured. Furthermore, in-depth data also gave insight into powerbalance in the relationship. Hereby, follow-up interviews could be based on experienceswhich took the investigation further.

Methodological cons in the RFID implementation studyDuring the research process several challenges arose. In the following paragraphsthese will be reflected on.

As a consequence, of close cooperation, a risk of a lack of distance between theresearcher and the researched group may appear. At best, this is a major advantage as“the true” picture is revealed. However, the current study also indicated that it is easyto adopt an internal perspective without critically examining taken-for-grantedinformation and accepting it as fact. Put another way, a researcher may adopt theblindness to defects of the researched object and thus lack an outside perspectivewhich often is claimed to result in fresh inputs.

Being an active part in the process investigated, it may be difficult to setthe boundaries between participating and observing. It was a balance to decide to whatextent the researcher should affect the process and still not conduct action research.For instance, with the author being one of two participants in an experiment, where onecrucial element was to brainstorm about how to solve problems, the line between

IJPDLM37,2

158

participant observation and action research was rather thin. This issue was dealt withby actively discussing problems, but still letting the company representative come tothe final decisions. However, the boundaries between participating and observing mustbe reflected on in every single case and the choice of boundary is of a subjective nature.

It may be difficult to question the researched group’s behavior without affecting it.The researcher needs to scrutinize the researched group, but critical questions mightaffect the group’s forthcoming behavior. This might not necessarily be wrong, butneeds to be taken into consideration when reporting the investigation. It was a delicateand tricky issue to analyze.

Six pragmatic challenges, previously identified by Jackson (1983), are likely toappear in participant observation. In the RFID study, these challenges were tackled invarious ways. First, the extensive amount of data was sorted and analyzed with a greatdeal of effort and reflection. Second, to avoid the negative impact of a time gap betweenthe occurrence of the event and its being recorded as data, which might result in vitaldetails being omitted, the author tried to take detailed notes during the whole study.Third, to minimize the risk of the negative influence of a lengthy delay betweenresearch and writing, recording and analysis of data were carried out continuouslyduring the investigation. Furthermore, follow-up interviews were conducted afterparticipant observation. Fourth, the difficulties of analyzing infinite amounts of datawere overcome by systematic analysis using analysis techniques inspired by groundedtheory (Glaser and Strauss, 1967; Strauss and Corbin, 1998), for example, coding andcontinuous comparison. Fifth, to try to avoid bias from key informants, triangulationwas conducted. Thus, several data sources for similar events were sought, for instance,follow-up interviews of observed findings, or documents and observations whichprovided similar results. Finally, ethical challenges, such as, how to report disputes,lack of management support, and protect against intrusion on personal privacy, weredealt with by the author. He explained the intentions of the study to the observed groupand assured them that their anonymity was guaranteed in the report.

Participation vs observationThe degree of participation, and consequently the effect of a researcher may vary indifferent parts of an investigation. This was relevant in the progress of the projectstudied. The study visits were thoroughly objective as they mainly comprisedobservation, while the experiment was fairly subjective because participation wasemphasized. The working group meetings were somewhere in between. The author’smain stance was to observe, but his role was also to contribute with knowledge whichmay have moved the project in certain directions.

However, to be able to further reflect on the implication of degree ofparticipation/observation in the different data collection methods used in the project,the meaning of subjective and objective needs to be clarified. According to LongmanDictionary of Contemporary English (2005), “a statement, report, attitude, etc. that issubjective is influenced by personal opinion and can therefore be unfair.” Objective, onthe other hand, is “based on facts, or making a decision that is based on facts ratherthan on your feelings or beliefs.”

This discrepancy between the characteristics of subjective and objective, which tosome extent is associated with participation and observation, did not lead to obviousdifficulties in the study but may have affected the outcome. Theoretically, the

Participantobservation in

logistics research

159

researcher could have had a less active participative role in the experiment, but due topragmatic reasons this was not feasible because the researcher was needed as aresource. Furthermore, the interaction between researcher and practitioner in theexperiment was fruitful and accordingly the participatory approach probablyimproved the outcome. The interaction certainly increased the researcher’sunderstanding of the project.

It was found that access could be divided into two types. First, in the preparationphase there was a rather formal basic access which means obtaining permission toconduct the study and then collect data. The second type is slightly more difficultto describe, but we can call it fresh access. It was more informal and needed to bemaintained on a regular basis in the data collection phase. It may be assured bygaining trust and continuously showing contribution. In the working group theresearcher was involved in discussions and tried to contribute with knowledge.Consequently, it was possible to ask questions, which also facilitated the researcher’sunderstanding. Additionally, the author found that active engagement in the projectresulted in greater trust from the group and thus more access was gained.

Discussion and conclusionThe interorganizational RFID implementation studyEven though relationships are commonly studied in logistics and that SCM has receivedincreased attention, a recent study reveals that there are very few interorganizationalstudies and that “the current research has failed to look at that perspective of the SCM”(Sachan and Datta, 2005, p. 674). The RFID investigation discussed in this paperhighlighted opportunities to conduct research on interorganizational settings withparticipant observation, and it proved to be beneficial. Accordingly, it indicated that thismethod was suitable to investigate this essential perspective of SCM.

The fact that direct observation methods, such as, participant observation, appearto be rather uncommon in logistics research may contribute to the lack ofinterorganizational studies. Therefore, in order to fill the increased need forinterorganizational research, in the form of, for example, SCM and collaboration, theauthor suggests that more logistics researchers would benefit from consideringparticipant observation in their studies.

The interorganizational RFID implementation offered an example of the value andchallenges of participant observation in logistics. First, it demonstrated that accessmaybedifficult to obtain, but when this has been achieved a whole new world of detailedinformation may be discovered. Second, it presented the opportunity for researchers toacquire an in-depth understanding of a studied phenomenon, which in this case was aninterorganizational relationship in an implementation project. Third, the in-depthinformation gained served as a frame of reference, which facilitated reflection. Fourth, thestudy identified a challenge regardinghow to avoid internal prejudices.The challenge is tobe able to take a step back. Fifth, this challenge is connected to the previous point andconsidereda difficulty in settingboundaries betweenparticipating andobserving. Sixth, itwas also found that investigating the group without affecting it too much was a delicatebalance to achieve. Finally, the paper identified a number of pragmatic challenges in datacollection and handling which have been previously reported.

Based on the RFID implementation study presented in this paper, a proposition isthat more logistics research would benefit from participant observation with a slight

IJPDLM37,2

160

influence from ethnography. This view supports Naslund (2002, p. 332), who putsemphasis on the need for logistics researchers “to gain extreme relevance by spendingmore time in organizations . . . by ‘hanging out.’” Thus, the currently overlookedinterorganizational perspective in SCM would be more likely to be researched.

The opportunities presented by participant observation are also found elsewhere.Ellram (1996, Table 3) highlighted the usefulness of this method in qualitativeresearch. She claimed that it could be used for research aiming to explore, explain,describe, or even predict. By “hanging out” in organizations, participant observationcombined with other methods could result in more thorough logistics studies.

The dominance of the positivistic paradigm in logistics research (Mentzer andKahn, 1995; Naslund, 2002; Nilsson, 2005) may affect how participant observationresearch is valued in the logistics field. Considering paradigm, Easton (1995, p. 381)claims that in positivism “cases are only useful as exploratory devices,” but for a realistone case is enough as realists are “not generalizing to any population but to a realworld that has been discovered.” These quotes indicate two essential issues. First, itmight be challenging to generalize from a participant observation study in logistics.Second, it might be important to reason about epistemological standpoints whenmaking an analytical generalization in logistics.

In addition, Naslund (2002, p. 332) points out that:

. . . to some extent, ethnography is based on the rejection of positivism, and particularly therejection of the view that (social) research should adopt scientific methods consisting ofrigorous testing of hypotheses utilizing quantitative measurements.

This comment is valid for participant observation too, as it bears similarities toethnography. Obviously, rejecting positivism or at least keeping an open mind to otherepistemological positions to make use of participant observation might be a majorchallenge, but previous studies appear to call for such a stance. Moreover, the quotealso indicates that the dominance of positivism plays down the value of participantobservation research. Therefore, the author suggests that it may be essential to discussand argue for the significance of results obtained with this method. Another issue toexamine further is epistemological obstacles for participant observation in logistics.

In summary, this paper showed different views of participant observation intraditional logistics research (Ellram, 1996; Yin, 2003) and more socially focusedresearch (Atkinson et al., 2001; Hammersley and Atkinson, 1995). In addition, itpresented an overview of the characteristics of participant observation and comparedit with other qualitative methods. It also demonstrated the usefulness of participantobservation in logistics and particularly when investigating interorganizationalaspects. This usefulness was illustrated with an RFID implementation study, whichrevealed concrete methodological issues. The study also served as input to a moreoverall discussion of participant observation in logistics. Future research should reportexperiences from other participant observation studies regarding, for example,opportunities and obstacles, and to what extent a participant observation study shouldbe subjective and objective. It may also be directed at a deeper investigation of how toanalyze and report data from a participant observation study, and the role of thismethod in a case study. In addition, the brief analysis of epistemological aspects in thispaper needs to be extended.

Participantobservation in

logistics research

161

References

Alvesson, M. and Skoldberg, K. (1994), Tolkning och reflektion – Vetenskapsfilosofi och kvalitativmetod, Studentlitteratur, Lund.

Angeles, R. (2005), “RFID technologies: supply chain applications and implementation issues”,Information Systems Management, Vol. 22 No. 1, pp. 51-65.

Atkinson, P., Coffey, A., Delamont, S., Lofland, J. and Lofland, L. (2001), Handbook ofEthnography, Sage, London.

Bryman, A. (2002), Samhallsvetenskapliga metoder, Liber Ekonomi, Malmo.

Creswell, J.W. (1994), Research Design – Qualitative and Quantitative Approaches, Sage,Thousand Oaks, CA.

Dubois, A. and Gadde, L-E. (2000), “Supply strategy and network effects – purchasing behaviourin the construction industry”, European Journal of Purchasing & Supply Management,Vol. 6 Nos 3/4, pp. 207-15.

Easton, G. (1995), “Case research as a methodology for industrial networks; a realist apologia”,in Turnbull, P., Yorke, D. and Nande, P. (Eds), IMP 11th International ConferenceProceedings, pp. 368-91.

Ellram, L.M. (1996), “The use of case study method in logistic research”, Journal of BusinessLogistics, Vol. 17 No. 2, pp. 93-138.

Foote Whyte, W. (1991), Participatory Action Research, Sage, Thousand Oaks, CA.

French, J.R.P. and Raven, B.H. (1958), “Legitimate power, coercive power, and observability insocial influence”, Sociometry, Vol. 21 No. 2, pp. 83-97.

Glaser, B.G. and Strauss, A.L. (1967), The Discovery of Grounded Theory – Strategies forQualitative Research, Aldine Publishing Company, Chicago, IL.

Glaser, J.M. (1996), “The challenge of campaign watching: seven lessons of participant –observation research”, PS: Political Science and Politics, Vol. 29 No. 3, pp. 533-7.

Hammersley, M. and Atkinson, P. (1995), Ethnography – Principles in Practice, Routledge,Abingdon.

Jackson, P. (1983), “Principles and problems of participant observation”, Geografiska Annaler:Series B, Human Geography, Vol. 65 No. 1, pp. 39-46.

Jones, P., Clark-Hill, C., Shears, P., Comfort, D. and Hillier, D. (2004), “Radio frequencyidentification in the UK: opportunities and challenges”, International Journal of Retail &Distribution Management, Vol. 32 No. 3, pp. 164-71.

Karkkainen, M. (2003), “Increasing efficiency in the supply chain for short shelf life goods usingRFID tagging”, International Journal of Retail & Distribution Management, Vol. 31 No. 10,pp. 529-36.

Longman Dictionary of Contemporary English (2005), available at: www.ldoceonline.com(accessed 3 March 2006).

McFarlane, D. and Sheffi, Y. (2003), “The impact of automatic identification on supply chainoperations”, The International Journal of Logistics Management, Vol. 14 No. 1, pp. 1-17.

Mentzer, J.T. and Kahn, K.B. (1995), “A framework of logistics research”, Journal of BusinessLogistics, Vol. 16 No. 1, pp. 231-50.

Merriam, S.B. (1994), Fallstudien som forskningsmetod, Studentlitteratur, Lund.

Naslund, D. (2002), “Logistics needs qualitative research – especially action research”,International Journal of Physical Distribution & Logistics Management, Vol. 32 No. 5,pp. 321-38.

IJPDLM37,2

162

Nilsson, F. (2005), “Adaptive logistics – using complexity theory to facilitate increasedeffectiveness in logistics”, doctoral thesis, MediaTryck, Lund.

Palsson, H. (2006), “Implications of interorganizational RFID implementation – a case study”,Conference Proceedings of ELA 2005, Vol. 10, pp. 85-100.

Park, P. (1999), “People, knowledge, and change in participatory research”, ManagementLearning, Vol. 30 No. 2, pp. 141-57.

Prater, E., Frazier, G.V. and Reyes, P.M. (2005), “Future impacts of RFID on e-supply chains ingrocery retailing”, Supply Chain Management: An International Journal, Vol. 10 No. 2,pp. 134-42.

Ragin, C. (1992), “‘Casing’ and the process of social inquiry”, in Ragin, C.C. and Becker, H.S. (Eds),What is a Case? Exploring the Foundations of Social Inquiry, Cambridge University Press,Cambridge, pp. 217-26.

Rutner, S.M., Waller, M.A. and Mentzer, J.T. (2004), “A practical look at RFID”, Supply ChainManagement Review, Vol. 8 No. 1, pp. 36-41.

Sachan, A. and Datta, S. (2005), “Review of supply chain management and logistics research”,International Journal of Physical Distribution & Logistics Management, Vol. 35 No. 9,pp. 664-705.

Seuring, S. (2006), “The rigor of case study research in supply chain management”, in Persson, G.and Jahre, M. (Eds), Nofoma 2006, Oslo.

Sheffi, Y. (2004), “RFID and the innovation cycle”, International Journal of LogisticsManagement, Vol. 15 No. 1, pp. 1-10.

Smaros, J. and Holmstrom, J. (2000), “Viewpoint: reaching the consumer through e-grocery VMI”,International Journal of Retail & Distribution Management, Vol. 28 No. 2, pp. 55-61.

Strauss, A. and Corbin, J. (1998), Basics of Qualitative Research, Sage, Thousand Oaks, CA.

Thomsson, H. (2002), Reflexiva Intervjuer, Studentlitteratur, Lund.

Yin, R.K. (2003), Case Study Research: Design and Methods, Sage, Thousand Oaks, CA.

About the authorHenrik Palsson is a PhD Candidate in Logistics at Lund University, where he also received hisMSc in Mechanical Engineering. His doctoral thesis is focused on interorganizationalcollaboration in logistics. Prior to his PhD studies, he has worked at a logistics consultancy firmfor four years and at a manufacturing company for one year. Henrik has published in conferenceproceedings and he was awarded for best paper at the Nofoma Conference 2006. Henrik Palssoncan be contacted at: [email protected]

Participantobservation in

logistics research

163

To purchase reprints of this article please e-mail: [email protected] visit our web site for further details: www.emeraldinsight.com/reprints