In a research project aimed at suggesting improvements at an annual advent fair three different ethnographic research approaches were used; Social Anthropology, Interaction Design and Mobile Ethnography. The paper focuses on how the three different approaches on ethnography affected choices in the research process, the outcomes of the research and how the outcomes were presented. It is found that the different motivations for doing ethnography between the three approaches make their outcomes differ in a clear way. These differences make the three ethnographic approaches suitable for achieving different research outcomes.
From its early origins in the works of Boas, the Torres Strait island expedition, and Malinowski, the ethnographic approach has moved in many directions. One of the many disciplines which has adopted and appropriated ethnography is User-Centered Design (UCD). The various branches of UCD aim to create services, products and interactions based on insights about the “needs” and habits of the people who are the intended users of the artifact being designed – ethnography has established itself as a crucial tool to gain that understanding. Yet another development is the emergence of Mobile Ethnography which guides smartphone owners in collecting ethnographic data about themselves. These forms of ethnography however often differ from how it is conducted in academic anthropology.
This paper offers a comparison showing how different ethnographic approaches produce different kinds of knowledge while investigating the same empirical context and broad brief. The paper is based on a recent fieldwork experiment with separate research teams using Social Anthropology, Interaction Design and a Mobile Ethnography app respectively.
COMPARING STYLES OF ETHNOGRAPHY
Ethnography’s journey from Anthropology to design and the design community’s motivations for incorporating ethnography in its toolkit has been described in a number of publications throughout the years (e.g. Grudin, 1990; Ford & Wood, 1996; Segelström, Raijmakers, & Holmlid, 2009; Segelström, Holmlid, & Alm, 2009; van Dijk, 2010) and will not be discussed further in this paper. Nor will the argument that designers have misconstrued anthropological ethnography be explored any further than this acknowledgement of the (former) existence of such a discussion (see Dourish, 2006, for an example of this discussion). Rather, this short background section will focus on previous comparative studies of ethnographic styles.
At EPIC 2008, Dori Tunstall (2008) presented a comparison of how the basic approach to ethnography differs between anthropology, marketing and design. Her study was based on a literature review of how the three fields are described, grounded in one prototypical company for each approach. She used a framework for theory suggested by Barnard (2000), called QAME. QAME stands for questions, assumptions, methods and evidence. Tunstall explores each of the categories and maps how the three different approaches to ethnography differ. Table 1 summarizes her findings.
TABLE 1. Tunstall’s analysis of ethnographic approaches. Adapted from Tunstall (2008, p. 220).
Anthropologist Rob van Veggel (2005) published an account of potential clashes which can occur when anthropologists and designers collaborate, due to the differences in motivations for using ethnography. He bases his comparison on recollections of his experiences whilst working as an anthropologist for design companies. He explains how designers and anthropologists approach ethnography:
“[D]esigners approach ethnography for the practical reasons of gaining a rich and deep understanding of users that can be easily integrated into design projects, and yet quick and relatively inexpensive to obtain.” (van Veggel, 2005, p. 5)
“[A]nthropologists approach, ethnography as the methodological component of a theoretical endeavor to understand humans as socio-cultural beings, who presumably act and think in different way; ethnography is a method to understand other people – Anthropology is that understanding” (van Veggel, 2005, p. 8)
Based on these differences van Veggel (2005) identifies four potential conflicts when the two disciplines collaborate: the translation of anthropological insights into usable insights for designers, designers lacking in preparation and training before doing ethnographic work, how to link people’s actions and thoughts together and that Anthropology is excessively theoretical.
It is evident that both previous comparative studies presented see a clear difference between the design take on ethnography compared to the anthropological one. The motivations to use ethnography differ, leading to different aspects being seen as the important ones during the ethnographic process.
One of the goals of this paper is to assess the claims of van Veggel (2006) and Tunstall (2008) through an empirical study.
The underlying case for the analysis presented in this paper is a research study at an annual advent fair in Linköping, Sweden. The researchers had been invited to contribute with insights on the visitor’s experience of the fair, insights which could aid in the renewal of the fair offering in upcoming years.
The advent fair is held in a small neighborhood called Gamla Linköping – “Old Linköping” -, which also is an open air museum. The neighborhood was constructed in the 1950’s as Linköping, like most other Swedish towns, was modernized in terms of building standard (Gamla Linköping, n.d.). The neighborhood consists of houses originally built in other parts of Linköping, but rather than being torn down they were moved to Gamla Linköping. The neighborhood is meant to give its’ visitors a feeling of what a Swedish town looked like in the early 20th century. Some smaller adaptions to modern life have however been done, such as opening up a gravel path in the streets to ease the pushing of trolleys and replicas of old houses being built from scratch.
The advent fair is held every year on the first advent weekend in cooperation with the local Lions Club. Being held only two days a year, the fair introduced a strong temporal constraint on how the research could be conducted. However, the fair is a popular event with around 15000 people visiting it the study year, meaning that there was a large pool of potential study participants to draw from. Taking the time constraints and the unique environment of the fair into account it was decided to use several approaches to understand the visitor’s experience of the fair. Three different research approaches were used (Social Anthropology, Interaction Design and Mobile Ethnography), each with a brief corresponding to what could be expected had they been hired as consultants directly by the fair management. Each approach is described in further detail in the next section.
THE THREE ETHNOGRAPHIES
After the decision had been made to use multiple approaches, the researchers decided to set up a study which would produce data from different theoretical approaches. The researchers having access to the beta-version of a Mobile Ethnography application, it was decided to conduct an experiment where different ethnographic styles would be compared.
Mobile Ethnography is a still emerging method, which led to the decision to use two other approaches with varying degree of maturation. Returning to the roots of ethnography to use anthropologists was quickly decided upon as the most mature approach. Considering the authors work in UCD, and the Mobile Ethnography app being inspired by service design, the UCD approach to ethnography, in form of interaction design, was chosen as the mid-point maturity wise.
In the following sections the approaches used by the fieldworkers coming from the different theoretical starting points are expanded upon. We do not claim that the teams conducting the research for this project are necessarily ‘typical’ or ‘representative’ of how these approaches should be applied.
Two thesis students in Social Anthropology were recruited as fieldworkers. Their project brief was to do a study according to the tradition of Social Anthropology which could aid in the improvement of the visitor’s experience of the advent fair in coming years. The Social Anthropology head at Linköping University volunteered to be the advisor of the theses, (and to summarize their contents). The description of the approach of the social anthropologists below is based on his report (Alm, 2012) as well as the two finished theses (Karlsson, 2012; Nyman, 2012).
Both students followed a similar approach in their work. They started out by narrowing down the broad brief to match their research interests; in Hanna Karlsson’s case she focused on the shopping experience of her informants and Mikaela Nyman focused on Gamla Linköping as a reconstructed past. This was followed by a broad phase where they acquainted themselves with literature on their thesis topics, Gamla Linköping as a field work site and finding informants. Aided by insights from the literature studies both anthropologists conducted semi-structured interviews with their informants.
The interviews were then transcribed with the aim of finding reoccurring themes and verifying/falsifying the insights from the literature study. The emerging themes were used as a foundation for participatory observations of the informants’ visits to the advent fair.
Thereafter the anthropologists analyzed the material from their observations and compared it with the insights from the interviews. They then re-visited their informants with follow up questions to clarify issues which were unclear. The answers from the follow up questions were integrated into the rest of the analysis, producing the end result analysis wise.
The interaction designers in the project were chosen to match the anthropologists experience-wise. The six students enrolled in the final Interaction Design course given to master students in cognitive science were given the brief to develop concept ideas for interactive artifacts, based on the wishes and driving forces of the visitors to the advent fair. The user research and presentation of it was stressed as a key learning moment in the brief. As (interaction) designers usually work in teams, the group was divided into two teams. Each team was given free choice in planning their ethnographic work. This resulted in the two teams conducting somewhat different tasks.
One team decided to do benchmarking of other advent fairs, to use structured interviews to interview people at the fair and to recruit two groups of friends to interview prior and after their visit to the advent fair as well as to shadow them during their fair visits. One group was shadowed at the advent fair in Gamla Linköping, and another at a competing fair. As the group analyzed their material they did however mostly use the interviews, and the team even states that “we’ve only used a small section of the observation data” (author’s translation) in their project report. The questions asked focused on the motivations for visiting the fair, what was bought and how/if the informants searched for information about the fair prior to visiting it. The analysis and synthesis led the team to creating two personas1.
The other team started out by formulating hypotheses about what would be interesting aspects to study, formulating two questionnaires based on this. The focus was narrowed further by concentrating on two user groups; families with young children and retired people. The team had initially aimed at getting 30 questionnaires back from each group, but experienced difficulties in recruiting people on-site, only managing to get 29 questionnaires answered in total during the first day of the fair. The team thus decided to do undirected observations during the second day of the fair. Like the other team, they focused on the more tangible data, letting the observations take a backseat in the analysis. The analysis and synthesis of one of the questionnaires led them to create three user profiles, after having weighted the questions during the analysis. The team did not find any clear patterns in the second questionnaire.
The third research approach of the study was to use a smartphone application, in which the participants documented their visit to the advent fair without the researchers being present. The creators of the app have dubbed this approach Mobile Ethnography(Stickdorn, Frischhut, & Schmid, 2012): “[m]obile ethnography is a research approach to identify, evaluate and document the customer journey through a smartphone application” (Frischhut, Stickdorn, & Zehrer, 2012, p. 161).
Although not explicitly dealt with by the creators of the app used, this approach relates to ethnographic discussions on multi-sited ethnography (Marcus, 1995), autoethnography (Anderson, 2006, and comments; Solomon, 2010) and ethno-mining (Anderson, Nafus, Rattenbury, & Aippersbach, 2009).
myServiceFellow – The app myServiceFellow2 (mSF) is developed by a research group focusing on service design for tourism. The motivation for developing the app is that a holiday maker’s experience of the trip starts when the trip is planned and isn’t finished until they arrive at home again – tourism destinations need to understand this to be able to deliver a top-notch experience. Stickdorn, Frischhut & Schmid (2012) highlight the difficulties of getting this holistic view by only interacting with the holiday makers at the destination, as the holiday is a journey through many geographical places (the same kind of issues discussed around multi-sited ethnography).
The approach of the mSF-development team to issues of multi-sited ethnography is to let the holiday maker’s document their own holiday (relating both to autoethnography3 and the design probes approach (Mattelmäki, 2006)). The holiday makers are guided in documenting their holiday by a smartphone app, built according to the customer journey metaphor used in service design. The users of mSF add new data points, called touchpoints in the app, at their own discretion as note-worthy events occur:
“It is the guests who decide what is a touchpoint during their individual customer journeys and it is them who evaluate and document those by adding text messages, pictures, videos or audio files besides meta data such as date, time and GPS position” (Stickdorn, Frischhut, & Schmid, 2012, p. 7)
As shown in the quote above the app makes use of the functionality available in modern smartphones, such as allowing the users to create video and photo material (see Faulkner & Zafiroglu, 2010, for a discussion on benefits of user-generated video and photo material) as well as ethno-mining of the holiday maker’s movements.
Once the holiday maker has finished her holiday and uploaded the material, it becomes available to the project owners in the back-end of mSF, which is called ServiceFollow. All users’ materials are shown in a chronological order by default, together with their rating of the occasion (ratings go from -2 to +2). ServiceFollow provides the project owners numerous ways of interacting with the material; re-sorting, varying degrees of detail, accessing images and videos, grouping of touchpoints and a map view (based on GPS-position) among others.
Methodology – The mSF-study used a mixture of pre-recruited and on-site recruited participants. A total of 26 respondents uploaded data from the advent fair. Three support personnel were available at the advent fair to support the participants who needed assistance. This team also handled the on-site recruitment by approaching visitors as they entered Gamla Linköping.
The participants were informed that the insights gathered through the tool were going to be used to suggest improvements for upcoming advent fairs.
The focus of this paper being the similarities and differences between three approaches to ethnography, the findings section focuses on a comparison between them
. The analysis highlighted a number of areas in which there were differences between the approaches. These areas are presented in the order they would appear in the ethnographic process.
The benchmarking done by the anthropologists in the project focused on the existing literature on their topics. They spent longer time on the benchmarking efforts compared to the other two approaches, searching for and reading literature. The benchmarks were used to help them narrow down their focus and decide on what could be interesting to focus on in in their initial interviews. Furthermore the analysis of the collected material was aided by including insights from other anthropologists’ studies of similar topics.
The interaction designer team which did benchmark focused on how other advent fairs and Christmas markets have solved the issues they hypothesized would be interesting, such as way finding and event calendars. The benchmarking was used as inspiration during the ideation, but did not (explicitly) affect their own user research.
There is no built in benchmarking in the Mobile Ethnography app that was used, but the participants did their own benchmarking comparing with their expectations and previous experiences. A reoccurring example was that many users complained about the weather, as the winter weather during the market was the worst conceivable; windy nor any snow, whereas it normally would be snow and no wind worth mentioning.
Scope of study
The anthropologists started their projects by narrowing down their overall focus to specific research questions. Once the scope of the study was set it was not challenged again. The strong delimitation of the subject area however gives the opportunity to explore it in detail and have a deep research approach.
The interaction designers started by formulating a hypothesis on what could be interesting, which were used as starting points. The path set by the hypothesis could be left at any time if the emerging insights pointed to better design opportunities somewhere else. Thus new hypothesizes could emerge at any time. This means that the scope of the interaction designers is open. The openness however comes at the price of depth – the interaction designers’ study does not consider the phenomena studied at the same level of detail as the anthropologists.
The overall scope of the Mobile Ethnography study is set by the organizers of the study, and is manifested in how they brief the participants. However, there is no way of making certain that all participants stay true to the intended scope, they may misinterpret instructions or (unconsciously) delimit the scope further.
What is made the focus of the study?
Whereas the scope of the study relates to width and depth of the study, the focus of the study relates to which aspects are studied within the scope. Focus of the study closely corresponds to the ‘Questions’ heading in Tunstall (2008).
The anthropologists focused on behaviors and motivations, and how what is observed relates to the existing body of knowledge. Being able to describe a small section of human behavior is the goal of the study. In the long run this is a part of what can be called “the anthropological task” – to describe humanity through detailed studies of all aspects of human behavior (see van Veggel, 2005, quote above).
In contrast, the interaction designers focus on finding
s situations which currently do not live up to expectations (or which can meet the unmet needs of the users). As a part of this the constraints on potential design solutions are also of interest. The goal of the user research is to find design opportunities within the current situation.
In Mobile Ethnography which aspects get studied are once again in the hands of the study participants. The things which happen to happen, where the participants are, and their experience of events, become the focus of each individual’s participation (and data record). The overall focus for this approach thus emerges from the events which are most frequently chosen by the participant group as a whole.
Where and how are data collected?
The approaches on how to collect the data was similar between the anthropologists and interaction designers. Both used non-contextual interviews and contextual observations as their basic methodology. The difference between the two was mainly in the details; the anthropologists used a truly semi-structured approach to the interviews whereas the interaction designers had a structured approach with the possibility to add non-prepared follow-up questions. During the contextual observations the anthropologists did participatory observation whereas the interaction designers choose passive observations. Both did however observe continuously, getting a full view of their participants’ fair visits. One difference, however, was that one of the Interaction Design teams chose to use surveys.
The Mobile Ethnography approach stands in contrast to the other two; it was only done in context and at the participants’ discretion. As the participants added touchpoints only as they saw fit, the data gives ‘snapshots’ of the visit rather than a continuous description.
How is the analysis performed?
The analysis done by the anthropologists was thematically based, taking advantage of the different types of data collected and done without the help of any software. The differences between the various types of data collected are used to build a stronger argument for the conclusions. When the preliminary analysis was finished, the informants were approached again to fill in knowledge gaps.
The interaction designers seemed to view the analysis as a tool to support the synthesis. Both teams searched for patterns in their material through manual sorting and resorting, but mainly used the more tangible interview and survey data, neglecting their observations.
The Mobile Ethnography app mSF and its back-end ServiceFollow provides the project owner with the raw material submitted by the participants, thus making it the project owner’s responsibility to conduct the analysis (according to whatever methodology they choose to use). There is functionality built into ServiceFollow to ‘help’ the project owner with the analysis, but usage is optional.
How is the synthesis performed?
Synthesizing the material is not part of the anthropologists’ usual process; any synthesis which occurs is a by-product of the analysis.
On the other hand, the synthesis is the main focus for the interaction designers. They grouped and ordered their insights, aiding them in creating tangible evidences of their user research, manifested through the two teams’ creation of personas/user profiles. These tangible summaries of the user research are later used as both inspiration for the design work and as a validator of the suggested design ideas.
As with the analysis, myServiceFellow and ServiceFollow do not provides the project owner with a synthesis but offers tools to aid them in performing the synthesis.
What is more important – analysis or synthesis?
As the discussions on analysis and synthesis point towards, there is a difference in what is seen as the most important in the translation of raw data to insights (and reports); the anthropologists put their emphasis on the analysis whereas the interaction designers focus on the synthesis. ServiceFollow is neutral in this regard; having functionality to support both analysis and synthesis, but leaving the decision on how to do it up to the project owner.
Degree of interpretation
Degree of interpretation refers to how and to what degree the ethnographers transform the data during the analysis and synthesis. The anthropologists abstracted the material to a high degree, producing generalized insights. However, they stayed close to their informants and used them to illustrate their points even at the higher abstraction levels. For example, Karlsson (2012) found that her informants all reacted strongly against perceived knick-knack at the fair and wanted to have it removed from future fairs. However, all her informants had different opinions on what constituted knick-knack.
The insights produced by the interaction designers did not strive to abstract the findings into a larger context. Instead the insights were removed from the individuals who had participated in the research, trying to generalize in such a way that the main patterns could be highlighted in the form of personas/user profiles. The team using surveys even expressed disappointment in only finding patterns strong enough to create user profiles in one survey.
The interpretation offered by the back-end of myServiceFellow is that it automatically transforms the participants’ uploaded material into journeys of their visit. It thus keeps focus on every single individual and does not abstract their experiences in any way.
How are the findings presented?
The anthropologists relied on text to communicate their insights. The texts were descriptive and argumentative in their nature and made strong use of existing literature and the theories presented therein.
The material by the interaction designers also included text, but rather as background material to those interested in how their tangibles were created. Design concepts, use scenarios and personas/user profiles are the center of the reports produced by the two Interaction Design teams.
Finally, ServiceFollow presents the raw material sorted according to the customer journey metaphor used in service design. The tool offers functionality to organize the material in various ways, but the underlying journey metaphor is a constant in the tool. The project owner can however dive deeper into a data point and participant generated video/photo/audio-material is not more than one or two clicks away.
What is presented?
Closely related to how the findings are presented is the question of what is presented (a question also discussed in Tunstall, 2008). The anthropologists presented their insights and how they relate to existing knowledge, providing a holistic view of the findings according to the scope of the study. As already indicated in the last section, the interaction designers presented their design concepts and user research supporting them – they did so by making storyboard-styled presentations of how their design ideas would be used at future fairs, based on the problems they saw. ServiceFollow gave the project owner access to and data submitted by the research participants using the app.
Applicability of presented findings
The final reports presented by the anthropologist give a theoretical description of the situation as it is, not suggesting any changes. This is in line with the anthropological tradition, where descriptive accounts of the current state have been the standard outcome for a long time. To be able to use them for making design changes these theories need to be re-interpreted and adapted. Moreover, the re-interpreted theories are more applicable to long-term strategic changes to the fair than improving elements of the fair now.
The reports by the interaction designers are streamlined towards suggestions for design refinements. Their designs are almost directly applicable – some detail design might be needed before going into production. But the suggested designs are mainly add-ons to the existing fair or changes to specific components, such as augmented reality games, way finding apps and services to create tangible memories from the fair.
The applicability of the Mobile Ethnography data is a mixture of the qualities of the two other approaches; like Interaction Design it focuses on specific components of the market (both positive and negative), which are seen as the most important by the fair visitors. And like Anthropology the material is not directly applicable, but needs to be interpreted and used as input for ideation.
If the processes and outcomes of the three ethnographies conducted at the advent fair are compared with the two previous comparisons between different styles of ethnography presented in the background section (van Veggel, 2006; Tunstall, 2008), the similarities between the findings are striking. The suggestions made by van Veggel and Tunstall are confirmed empirically by the insights obtained in this study, where content overlaps; there is a match between Tunstall’s arguments in regard to the questions and evidence of Anthropology and Design (described here as focus of study and what is presented how). There is also a fit with van Veggel’s descriptions of the two fields which finds resonance here in our discussion of the focus of study and applicability of presented findings in this paper. For example, we saw that the Interaction Design students had trouble connecting people’s actions and thoughts together, and opted to mainly focus their fieldwork analysis on what was said.
Comparing the three approaches in the study, it is evident that the Mobile Ethnography approach is the outlier of the approaches. A question arising from the material is whether Mobile Ethnography (in the form of the myServiceFellow ecosystem) truly is a form of ethnography or whether it should be seen as something else. This in turn raises the counter-question how to delimit ethnography. Agar (1996, pp. 241-246) suggests the use of a student-child-apprentice learning role and the search for patterns as defining qualities for ethnographies. As our analysis points out, the back-end does not provide much theoretical or methodological guidelines for analysis or synthesis (search for patterns) of the material received from the participants. Nor does it provide a true apprentice role (which relates closely to the aim of creating empathy held forward in design (e.g. in Segelström, Raijmakers & Holmlid, 2009). Thus, the beta-version of myServiceFellow used in the study does not pass current delimitations of what a full ethnographic approach is. However, a skilled user of myServiceFellow could end up with results which are close to fulfilling Agar’s criteria for what constitutes ethnography, aided by their previous knowledge in fields like Social Anthropology or Interaction Design.
Differences between the three approaches
Comparing the three approaches to ethnography it is evident that they differ from each other in many ways. These differences can be explained by the theoretical starting points used by the field workers. The differences manifest themselves in many ways, such as in the scope of the study, how the material is interpreted and to which degree.
In the case of Social Anthropology, they want to describe behaviors and make mental models apparent based on a very specific focus of the study in comparison to the two other approaches. This narrow focus made it possible for the anthropologists to go deep into the worlds of their informants and to get an understanding which was abstracted from their informants while still maintaining each informants perspective on the issues discussed. The focus of the work post-data collection for this team was to make the analysis true to the informant’s voices and produce an ethnographic text where the informants and their opinions were placed in a larger, descriptive, context. This deliverable is in line with the anthropological tradition.
Looking at the interaction designers we can similarly see how the overarching aim of their fieldwork (guided by the theoretical commitments of Interaction Design) affected their methodological choices throughout the project. The ethnographic work was directed at getting actionable insights; the initial focus was open to change at any time if the fieldwork pointed to more promising areas for future design work. This openness meant that the studies did not go as deep into any single aspect of the visitor’s experience of the fair as the anthropological study. Yet, on the other hand the interaction designers studied more aspects of the fair. The post-data collection work likewise focused on getting actionable insights for design, which meant that synthesis became the main focus in the processing of the material, leading to individual informants disappearing from the material (and the commonalities between various types of fair visitors being highlighted). The synthesis is however not seen as an end-goal, it is used as the inspiration for the design work, and later on as the guarantee for the relevance of the design solutions.
Finally, the Mobile Ethnography app myServiceFellow brings its own assumptions about the nature of the object of study, namely that it has a strong temporal aspect to it. As touchpoints cannot be reorganized by the individual participants, the phone app presumes that the order in which events happen is the most relevant to the participant. This assumption follows the material as it is transferred from the participant’s phone to the server of the project owners, who are also presented the data according to the timeline. The back-end offers the opportunity to re-sort the material but not to break free from the journey metaphor used along the timeline.
CONCLUSION: WHICH ETHNOGRAPHY WHEN?
Given the differences between the ethnographic approaches produced in this study the question of their relative qualities naturally arises. The answer naturally depends on the reasons for commissioning ethnographic work. In the case at hand, with a short time frame for fieldwork and improvement agenda the Interaction Design approach and its’ focus on actionable insights was the most helpful. However, as pointed out by Norman (2010), a user-centered design approach to ethnography is more likely to produce improvements to the existing situation, rather than bold transformation.
Keeping the focus on doing ethnography for applied purposes, the depth of the (social) anthropological approach provides insights which can drive long-term innovative change processes. The ethnography produced provides a rich description of the topics at hand, but without offering suggestions for how these descriptions can be lead to changes which will improve visitor satisfaction of the fair.
The Mobile Ethnography approach shows a lot of potential for ethnographic work where the object of study is more permanent than a weekend long fair. The material gathered by the visitors could provide great context to anthropological as well as design ethnographies which have a change agenda. By using myServiceFellow or similar apps for benchmarking, ethnographers can improve their possibilities to find the most rewarding focus for their ethnographies from their employers’ perspective. The Interaction Design approach would be aided in so far that it can remove some of the openness in the approach and be able to go deeper into the informants’ worlds whereas Social Anthropologists will be aided in specifying areas of studies which can produce more applicable ethnographies. We believe that the gap between ethnography for Interaction Design and Social Anthropological approaches will be lessened with the help of Mobile Ethnography. The study shows that to use a mixture of the approaches, if possible, gives the broadest amount of actionable insights.
To use an example based on insights from the study; several participants who used myServiceFellow commented on the number of people at the fair. Depending on the time of their visit some thought it was too crowded and some too few. Closely related, one of the social anthropologists (Karlsson, 2012) noticed that her informants avoided overcrowded market stands and found support in previous anthropological work on the amount of personal space needed to make people feel comfortable in shopping. Put together this gives us insights into the importance of trying to “even out” the amount of people visiting the fair at any given time to increase both the enjoyment of the fair for the visitors, and the likelihood of them buying anything from the stands. We also noticed that the most appreciated part of the advent fair – in the data retrieved via myServiceFellow – were mini-events (like dancing around the Christmas tree). Furthermore with the help of the time-stamps from the touchpoint data we could actually confirm that there was higher attendance at the fair during the hours where events were scheduled. By adding more events, or spreading out the existing events over time, we could thus improve both the visitor experience and the stand holder’s income. Concepts coming from the interaction designers included suggestions for how to make people more aware of scheduled events, based on their learning that most people did not know in advance when events were due to be held (even though these were a highlight of the advent fair for them).
In conclusion, this paper has described a study on how three (self-described) styles of ethnography led to different types of findings whilst studying the same advent fair. The study results give empirical support to previous papers by van Veggel (2006) and Tunstall (2008). Furthermore, the study shows how the theoretical starting points of the three ethnographic styles affect what will be presented as the outcome, and how that outcome is achieved.
The research presented in this paper was made possible through funding by European Union research project “Service Design as an approach to foster competitiveness and sustainability of European tourism”. The paper has benefitted through insightful comments from Stokes Jones, Mattias Arvola and the anonymous reviewers. Finally the authors wish to thank Hanna Karlsson, Mikaela Nyman, Olof Jönsson, Thea Dahlqvist, Jeanette Bendelin, Charlotte Isaksson, Daniel Ros and Sofia Klasson for their efforts in the project as well as Björn Alm for mentoring the social anthropologists.
Fabian Segelström is currently pursuing his PhD at Linköping University, working in the Interaction and Service Design research group. His research focus is on how designers create an understanding for those they design for, and how they use and communicate this understanding once it has been achieved.
Stefan Holmlid is associate professor at Linkoping University and heads the Interaction and Service Design research group. He is a cofounder of the International Service Design Network and the Service Design and Service Innovation conference, ServDes. Currently his research circles around the expressive and inclusive powers of design methods and techniques in the service sector.
1 Personas/user profiles are a popular tool in design to create a number of prototypical users of whatever is being designed. They are used as an inspirational summary of research.
2 A closed beta version was used in this study.
3 The exact meaning of the term is autoethnography is still being discussed in the ethnography community. For the purposes of this paper it is understood as any effort to collect ethnographic data where no external ethnographer is present. Also see Solomon (2010) for a comment on this discussion.
Agar, Michael H.
1996 The professional stranger: an informal introduction to ethnography, 2nd edition. USA: Academic Press.
2012 The Advent Fair in Gamla Linköping, 2011: A report on two student theses. Project report, Linköping, Sweden. Retrieved from www.ida.liu.se/~fabse/alm-adventfair.docx
Anderson, Ken, Nafus, Dawn, Rattenbury, Tye, & Aippersbach, Ryan
2009 Numbers Have Qualities Too: Experiences with Ethno-Mining. EPIC 2009 Proceedings (pp. 123-140). American Anhtropological Association.
2006 Analytical Autoethnography. Journal of Contemporary Ethnography, 35(4), 373-395.
2000 History and Theory in Anthropology. New York: Cambridge University Press.
2006 Implications for Design. CHI 2006. Montreal, Canada: ACM.
Faulkner, Susan A., & Zafiroglu, Alexandra C.
2010 The Power of Participant-Made Videos: intimacy and engagement with corporate ethnographic videos. EPIC 2010 Proceedings (pp. 113-121). Tokyo, Japan: American Anthropological Association.
Ford, John M., & Wood, Larry E.
1996 An overview of Ethnography and Systems Design. In D. Wixon, & J. Ramey (Eds.), Field Methods Casebook for Software Design (pp. 269-282). USA: John Wiley & Sons.
Frischhut, Birgit, Stickdorn, Marc, & Zehrer, Anita
2012 Mobile ethnography as a new research tool for customer-driven destination management – A case study of applied service design in St. Anton, Austria. CAUTHE 2012: The new golden age of tourism and hospitality; Book 2; Proceedings of the 22nd Annual Conference, (pp. 161-167). Melbourne, Australia.
n.d Mer om oss > Historik. Retrieved June 29, 2012, from Gamla Linköping: http://www.gamlalinkoping.info/index.php/sv/mer-om-oss/historik
1990 The Computer Reaches Out: The Historical Continuity of Interface Design. Proceedings of Computer-Human Interaction, CHI, 1990 (pp. 261-268). Seattle, WA, USA: ACM.
2012 Shop until you drop: En studie i konsumentbeteende och julmarknadens betydelse. Linköping: Linköping University Electronic Press. Retrieved from http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-74412
Marcus, George E.
1995 Ethnography in/of the World System: The Emergence of Multi-Sited Ethnography. Annual Review of Anthropology, 24, 95-117.
2006 Design Probes. Helsinki: University of Art and Design.
2010 Technology first, needs last: the research-product gulf. interactions, 17(2), 38-42.
2012 Mysiga Gamla Linköping: Det konstruerade kulturarvets historieanspråk och dess autencitet. Linköping, Sweden: Linköping University Electronic Press. Retrieved from http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-75162
Segelström, Fabian, Holmlid, Stefan, & Alm, Björn
2009 Back to the Roots: An New Ideal for Ethnographic Research for Interface Design. Proceedings of the International Association of Societies of Design Research, IASDR 2009. Seoul.
Segelström, Fabian, Raijmakers, Bas, & Holmlid, Stefan
2009 Thinking and Doing Ethnography in Service Design. Proceedings of the International Association of Societies of Design Research, IASDR 2009. Seoul.
2010 Becoming the Subject: A Comparison of Ethnographic and Autoethnographic Data for New Product Development. EPIC 2010 Proceedings (pp. 66-78). Tokyo, Japan: American Anthropological Association.
Stickdorn, Marc, Frischhut, Birgit, & Schmid, Josef Sebastian
2012 Mobile Ethnography as a Pioneering Research Approach for Customer-Centred Destination Management in European Tourism. Travel and Tourism Research Association 2012. Virginia Beach, VA: in print.
2008 The QAME of Trans-disciplinary Ethnography: Making Visible Disciplinary Theories of Ethnographic Praxis as Boundary Object. EPIC 2008 Proceedings (pp. 218-233). American Anthropological Association.
van Dijk, Geke
2010 Design Ethnography: Taking inspiration from everyday life. In M. Stickdorn, & J. Schneider, This is Service Design Thinking (pp. 108-115). Amsterdam, The Netherlands: Bis Publishers.
van Veggel, Rob J. F. M.
2005 Where the Two Sides of Ethnography Collide. Design Issues, 21(3), 3-16.