From “Thick” to “Broad” data?

So far, we haven’t dwelled too much on the cons of the way we approached ethnographic data collection in our case study of the political festival The People’s Meeting. But surely, as our ethnographic focus was prescribed by observation guides and each note typed in a semi-fixed template, there are caveats to consider. Like in more traditional quantitative approaches that we know from the natural sciences, we adopted a much more rigorous mindset than we were used to. With our toolbox of observation guides, the Ethno-platform, and other self-developed schemes, our data collection was preplanned in detail. Before entering the field, we had pinned down exactly what to observe, when to observe it, and how to take note of it. The downside to this approach was the little room left for each of us to pursue new paths or clues unfolding before our eyes. Paths which could not necessarily be predicted in our preplanning. Our approach to the field was far less explorative and flexible than traditional ethnographic fieldwork, putting our project at risk of missing themes which could be central to the analysis of attention flows at the festival. We consequently found that the price of systematization and rigorousness in this case became a tradeoff where we – to some extent – had to say goodbye to elaborate, thick descriptions from the field in favor of a comparable and computationally processable kind of ethnographic data.

Broadband Ethnography

While the data we collected clearly diverged from the more traditional thick ethnographic descriptions, we strived to obtain ethnographic insights which could contribute further than with context to the project. Instead, we collected what we term “broad” data which held compatibility with other data types as well as stand-alone quality. The broad character of our data comes from the three Cs; Compiling, Comparing, and Computational processing, described in the previous posts. Importantly, these qualities implied that the ethnographic data we collected would be compatible with other sorts of data collected at the People’s Meeting by team members on the project.

In telecommunication, broadband means fast transmission of multiple signals at a range of different frequencies. In the same way, we like to think of broad ethnography, or should we say broadband ethnography, as an approach that aims at collecting data which can easily be connected to a wide range of data types from different disciplines. We experienced that broad ethnography was highly useful in an interdisciplinary setting. The data we gathered were carefully filtered and collected with a clear analytical focus. The different empirical material seemed slimmed down on its own, but in combination, our data offered a broad coverage of attention dynamics at the People’s Meeting.

Utilizing Broad Data

An example of where the broadness of our data could come in handy was in combination with ticket sales data and Twitter data related to events at the festival. From the ticket sales data, we could get a sense of which events attract the attention of audiences prior to the festival, and then we could extract information from attention schemes and fieldnotes written during the same events to get a sense of the temporality of aspects of attention flows surrounding particular events. As for the Twitter data, it could be used to examine how political attention surrounding the festival also flows online. We found that stakeholders often tweeted about the issues raised at events at different times during the festival. Some tweets had more interactions in terms of retweets and comments. We were able use data from the Ethno-platform to examine whether certain issues received attention at the same time on the physical festival site as on Twitter by cross-referencing timestamps of tweets and fieldnotes. And by finding the corresponding attention schemes, we could also get a sense of the audience’s attention during the relevant events. In this way, we were able to zoom in and out of our ethnographic data while combining it with other data types. This meant that we could shed light on the attention dynamics at a political festival from several different angles at once.

Picture 9. Ethnographer in the field

Anthropological Ingenuity

Beforehand, we were not used to thinking of ethnographic data as something that could be compatible with other data types to the extent that we intended in this project, and we found ourselves on shaky ground when we started experimenting with computational processing of our broad data. The computational approach to analysis causes a risk of feeling loss of control as it involves handing over some important choices to the machine. Throughout the project, we strived to reach a balance where computational processing of fieldnotes and a structured approach to data collection could help align the data and contribute with analytical insights while keeping the anthropologist in the driver’s seat. Afterall, the depth of the anthropologist’s insight to a given field is one of the discipline’s finest strengths, and during our experiment, we found it useful to keep some flexibility left to qualitatively go through fieldnotes.

Moreover, we found that it was an advantage for us that the ethnographers were anthropologically trained as it took an experienced ethnographic eye to capture the most important dynamics in between the more quantitative observations. Afterall, the devil lies in the detail, and we believe that the nuances captured through the ethnographic observations were critical when observing attention flows at the People’s Meeting as part of an interdisciplinary project.

Conclusion

In this series, we have introduced how we have used a “broad” ethnographic methodology to a case study of attention at the political festival The People’s Meeting. In the beginning of the series, we stated that our approach would result in the collection of broad data due to the qualities of the three Cs which entailed that the data can be: Compiled, Compared, and Computationally processed. The data needed to be compatible with other, more quantitative data sources as we were part of the interdisciplinary project, DISTRACT. There are certainly discoveries that remain in the dark when approaching a field site with this sort of rigorous methodology, but it was a trade-off we were willing to accommodate in this specific study. We acknowledge that this approach isn’t suitable for all enquiries, and we certainly don’t wish to root out traditional ethnographic fieldwork in which we have great faith. We temporarily waved goodbye to the qualities of thick data and dug into the possibilities that broad ethnography might offer.

DISTRACT

The authors if this blogpost series are Sofie Læbo Astrupgaard — PhD fellow in Social Data Science at the University of Copenhagen, Clara Rosa Sandbye — PhD fellow at the Department of Anthropology at Aarhus University, and Emilie Gregersen — MSc student in Social Data Science at the University of Copenhagen. The trio has been working as a part of the interdisciplinary research project DISTRACT, studying the dynamics of issue attention at a political festival.

Source

Computational Processing of Ethnographic Data

After a few intensive days in the field, you and your team have returned to the familiar settings of the university. In front of you, there is a big pile of observation schemes and seating charts from the field awaiting you to turn them all into one common spreadsheet. Luckily, most ethnographers appear to have carefully recorded audience attention in full accordance with the instructions. After having typed in the last crinkled seating chart, you finally have a full overview of all the recorded quantified attention behavior from the field. You log on to the Ethno-platform to fetch a file with all the fieldnotes from the festival and load it to a programming application. You swiftly extract all the notes that accompany your newly created spreadsheet. Overwhelmed by this huge corpus of fieldnotes and observations, you wonder: Which computational techniques would be most helpful to find patterns in these data?

Picture 6. Structuring ethnographic data

Computational Potential

Computational programming is, unfortunately, often presented as more complicated or math-demanding than it needs to be. In many ways, it is like learning the grammar structure of a new language. As soon as you know the basic rules for how to construct a sentence and bend your verbs, you can slowly begin to communicate. Same thing with programming languages; when you understand the syntax and learn the basic logic behind building up a “script”, you can execute simple code. And even with a few basic skills, you can benefit from programming tools when working with ethnographic data. In the field of social data science, there have been different suggestions to how computers can help process and analyze ethnographic text: some find the machines helpful when coding their material; some have entrusted them with the responsibility to automatically code large parts of their fieldnotes; while others have used text mining techniques to explore notes and interviews to find new themes or patterns that they hadn’t noticed before. These are just a few examples of how computational potential paves the way for new ways to analyze ethnographic data. So, how did we put computational power to good use? We wanted to use computational techniques for two things: to explore our ethnographic material and to combine it with other data types that we collected at the People’s Meeting.

Uniting Ethnographic Data Sources

During the few days the festival lasted, we compiled a ton of beautifully aligned fieldnotes. When accessing the Ethno-platform, the infrastructure allowed us simply to press a button to fetch a file that contained all of them. We loaded the file to a programming application and converted it to a spreadsheet. Now they were ready for computational processing. Imagine a spreadsheet where each row holds the data of a fieldnote, and the different columns help to divide the different information and metadata related to that fieldnote (see Picture 7). Now, returning to the common format of our fieldnotes: each note was written, following three formalities (see Post 2) and holds meta-data about the described situation. Therefore, we could extract information by using these features with different search commands in the programming application. This meant that we could sort the data by date, time of day, ethnographer, event tent, and we could fetch quotes and analytical comments. These can surely be helpful features for the initial data exploration, and our aspirations to computationally process our fieldnotes were slowly being realized. However, we also wanted to combine our systematized quantitative observations with the spreadsheet of fieldnotes.

As alluded to in the beginning of this post, we turned the piles of attention schemes and seating charts into a common spreadsheet. The next step was to merge it with our fieldnotes from the Ethno-platform. The result was one grand spreadsheet of all our ethnographic data. The columns contained the text from fieldnotes and metadata as well as different levels of attention and seating information at each event. And though the data from the seating charts and attention schemes were of a different kind, namely reduced quantitative measures of attention and presence, they were now merged with the accompanying descriptive (though also structured) fieldnotes in which our group of ethnographers had strived to capture attention dynamics in interactions during events. We were now finally piecing together the somewhat fragmented ethnographic puzzle.

Picture 7. A spreadsheet of fieldnotes from the Ethno-platform

From Potential to Beneficial

With the united ethnographic data, we could finally begin to experiment with computational techniques for analysis. After having discussed different ways we could approach this sort of dataset, we decided to start simply by visualizing the quantitative observations of attention. In Figure 8, we have plotted the audience’s attention levels for each observed event. From our master spreadsheet, we extracted all events observed on Friday at the People’s Meeting (vertical axis). We used our observations of how many looked at their phone and at the stage throughout events to create a combined attention score for the two types behavior ranging from 0-10 for each 15 minutes of the event (horizontal axis). As each event lasted an hour this meant that the maximum attention score for an entire event is 40.

Figure 8 might not look very interesting at first glance but visualizing ethnographic observations does bring potential: it can guide parts of our analysis and bring some transparency to analytical choices. Questions and surprises emerging from what we see in the visualization of attention during events could be explored more by diving into the related fieldnotes. We can for instance examine how the audience preserves attention over time in a political event, and we can hold this up against the theme discussed during the event and observations from fieldnotes.[1]

From the visualization, we could also see that the approximate fraction of people paying attention to the stage was relatively stable overall across time intervals and across events, but we saw some small variations. And if we dove into the fieldnotes, we learned for the event with the lowest score, that it was extremely hot around the stage where the event was held. This meant that many in the audience were struggling with the heat, and instead of looking at the stage some were fanning themselves with magazines while others were focusing on ice cream they had bought before the event started.

Picture 8. Sorted bar chart of attention over time among audiences at different events

This was just one example of how we could explore our ethnographic data computationally. A possible next step could be to examine the differences between attention in the back and the front of the audience section, or to try to track temporal and spatial variation in attention at the festival site. When we had metadata recorded such as time and place for observations then we can also move on to merge other spreadsheets with other data types to our grand spreadsheet of ethnographic data. This could be data containing ticket sales for each event, tweets posted by event organizers, or maybe even weather data for each day during the festival. When we have the ethnographic data and metadata united in one spreadsheet loaded into a programming application then we can combine it with other data types.

So now we’ve unfolded our methodological and to some extent experimental approach to ethnographic data collection in an interdisciplinary setting. In the coming post, we will move on to discuss thick versus broad data and the implications of the kind of data we ended up collecting.

Notes

[1] We could for instance see that some ethnographers didn’t record attention scores all four times during events, as bars were missing for some events. In the fieldnotes from these events, we learned that this is due to events starting or ending early or the ethnographer arriving late.

DISTRACT

The authors if this blogpost series are Sofie Læbo Astrupgaard — PhD fellow in Social Data Science at the University of Copenhagen, Clara Rosa Sandbye — PhD fellow at the Department of Anthropology at Aarhus University, and Emilie Gregersen — MSc student in Social Data Science at the University of Copenhagen. The trio has been working as a part of the interdisciplinary research project DISTRACT, studying the dynamics of issue attention at a political festival.

Source

Producing Comparable Data through Systematic Observation

It’s 7:45 am. The morning briefing is about to start as you shove in the last bites of breakfast. One from the team is looking for the right cable to connect a laptop to the television screen. In a moment, the screen will display slides of today’s observation guide. Last-minute instructions are hurled out in the room as the clock strikes 8:30, and it’s time to go. Short of breath from hurrying to your designated event tent, you place yourself strategically, mobile phone in hand and ready to intensively observe and note down. With one eye on the clock and another on the audience, you note down anything of relevance in front of you. You alternate between counting the phone-scrollers, stage-watchers, and conversationalists every 15th minute and jotting down observations in the Ethno-platform. Overwhelmed by the many impressions, you wonder if you are following the detailed instructions like you’re supposed to. But exactly what measures are necessary to align our data collection for our purpose and make our observations comparable between researchers?

Picture 3. Morning briefing

Each morning, we held a briefing to make sure all ethnographers were in on the observation guide of the day (Picture 4). This was to ensure that everyone entered the event tents with the same analytical filter. When you venture into the craft of ethnography, you quickly realize that a million things happen at the same time. You cannot note everything down and all field observations are in principle an exclusion of other events you could have documented. By explicating exactly what we had to observe, we hoped to install a collective lens, which would capture the same type of attention-related behavior across researchers. The guides also served as a helpful tool to keep each of us on track of what to take notice of.

Picture 4. Example of observation guide

Quantified Attention-related Behavior

We figured that one way of streamlining our fieldnotes would be by counting attention-related behavior among the audience at different events. This idea of explicitly counting occurrences in the field is not very common in social anthropology where there seems to be a reluctancy to “mathematize” the discipline. However, some scholars hold that anthropologists do in fact count all the time in the sense that they register recurrences in the field to detect prevalent dynamics and themes. However, they rarely state the exact number of times a particular event happens. For our data collection at The People’s Meeting, we decided to deviate from traditional modes of doing fieldwork by explicitly quantifying ethnographic observations. This was done by developing what we call attention schemes and seating charts for the different event stages. The attention schemes and seating charts were distributed to each ethnographer alongside the observation guides at the morning briefing. Examples of these are shown in Picture 5 below.

Picture 5. Example of filled out observation scheme and seating chart

With the seating chart (Picture 5, right), we could spatially map where the audience were sitting at each event. In the example above, the ethnographer marked people present at the time that the event started with a black dot, and people arriving later with a ring. The attention schemes (Picture 5, left) were used to map attention behavior during the events. With the event tent divided into four sections (front-right, front-left, back-right, back-left), we noted if none (I), few (F), half (H), many (M), or everyone (A) were looking at the stage, at their phones, or talking to each other. We registered this in the attention schemes every 15 minutes.

Ensuring Comparability

Aside from the schemes and charts, we observed what took place in front of us in between the 15-minute intervals and wrote descriptive fieldnotes in the Ethno-platform. These tasks demanded our undivided attention if we were to uphold rigor in our data collection. Indeed, we find that this combination of observation guides, schemes, and the common format for fieldnotes provided by the Ethno-platform provided us with data that work well in combination. What we got was detailed records of how a given event progressed and different measures of the audience’s attention. And since we repeated the same procedure at each observed event, we can align and compare the data and hereby confirm or dismiss different tendencies we’ve observed across events and ethnographers throughout the festival site.

During the People’s Meeting we ended up collecting a ton of fieldnotes, seating charts, and attention schemes, and when we returned to the university it was time to reach our final goal of this project, namely, to process and analyze the data computationally – but more on that in the following blogpost.

DISTRACT

The authors if this blogpost series are Sofie Læbo Astrupgaard — PhD fellow in Social Data Science at the University of Copenhagen, Clara Rosa Sandbye — PhD fellow at the Department of Anthropology at Aarhus University, and Emilie Gregersen — MSc student in Social Data Science at the University of Copenhagen. The trio has been working as a part of the interdisciplinary research project DISTRACT, studying the dynamics of issue attention at a political festival.

Source

Compiling Ethnographic Data in an Ethno-platform 

The sun disappears behind the colorful town houses as you enter a pub in the narrow, cobbled road to test the pilot version of the Ethno-platform, an online fieldnote tool. It’s Wednesday, and tomorrow the town will buzz with people debating, networking, and navigating between each other around the festival area. You enter the pub and look for a table where you can sit discreetly, but still have a good overview of everyone in the pub. A group of friends immediately catches your attention. You begin to scribble: “20:58. They turn their chairs and move closer to a big TV screen beside their table. A UEFA match is about to start.” Just before, they were chatting eagerly with each other and now their common attention is oriented towards the TV. An analytical thought pops into your head, but you are unsure how to note it down in the platform. You wonder if it would be best to type it into the Ethno-platform next to the descriptive observation, or in a separate text field. Soon you return to the research team and discuss your experiences. How do we compile fieldnotes in a common format between researchers?

In our training as traditional (social) anthropologists, we’ve been told once and again that ethnographic work is a lonesome discipline conducted by a single ethnographer in the field. In the context of studying a comprehensive event like The People’s Meeting, however, the lonesome ethnographer might fall short. Since the festival only lasts for four days once every year, there was a great asset in mobilizing more ethnographers to cover more ground. We are sure that one trained ethnographer could collect rich data during the four days, but what if we could register what happens at every corner of the festival area at the same time?

The question remained, how we should go about this? Before us, sociologists and anthropologists at UCPH have experimented with what they call “short, big-scale fieldwork”. In the Utopia project, they asked themselves what kind of knowledge can be obtained if – instead of having one person conducting fieldwork in 100 days – 100 people conducted fieldwork in one day. Inspired by this idea, we asked ourselves how much data our team of ten scholars, of which seven were anthropologically trained, could collect in four days. In other words, we wanted to collect as many observations of micro-interactions at The People’s Meeting as possible. But clearly, this sort of collaborative data collection would require some coordination.

A Common Tool

A core task for the ethnographer is writing fieldnotes. This is often a messy and time-consuming process that entails jotting down in-situ notes in a notebook and elaborating on them later. The result is often unstructured and not easily comprehensible for anyone besides the author. We needed a way of streamlining our data collection to avoid a messy pool of observations without the time and resources to make sense of them. And after numerous considerations, this was how the first contours of the Ethno-platform emerged. So, what exactly is the Ethno-platform? The basic idea was to create a semi-fixed template for writing fieldnotes which would ease this sort of collaborative data collection[1]. Here, the ethnographers should be able to fill in their observations in pre-defined text fields on their device at hand such as a mobile phone or a tablet.

Picture 1. Ethno-platform interface on a mobile phone

Let us quickly guide you through the platform interface: When accessing the Ethno-platform on your phone, you meet the interface in Picture 1. The first thing you do is to select yourself as the authoring ethnographer from a list of team members on the project. Then you type in metadata such as date, location, and situation which is stored with the content of the fieldnote. In the “situation” field, you type in the necessary contextual information for other researchers to understand the observations described in the fieldnote such as “Panel debate on sustainable food industries”. Lastly, there are two open text fields. One, where you write your observations, and another, where you add analytical or methodological reflections to the set of observations or the project in general. In this way, all fieldnotes will have a similar structure while also allowing you to write descriptive notes and reflections from the field. When you click “done”, the notes are stored on a GDPR-compliant[2] server where they can be accessed and edited at any given point by all members of your team.

One of the main goals for the Ethno-platform was to make a common data archive where anyone from our project could access any fieldnote created during the week and in principle be able to utilize the data instantly. Here, the metadata from each fieldnote came in handy. Having consistently typed in the information for every fieldnote in the project during the case study, we ensured a simple contextual introduction to each note which helped everyone easily navigate in the fieldnotes through the platform.

A Common Format

Aside from the Ethno-platform aligning our fieldnotes in structure, we also needed to establish some ground rules for how the tool should be used in the field. This was key if we wanted to successfully collect numerous observations of the same type of micro-interactions. To do so, we agreed on three formalities when writing in the platform: Analytical or methodological comments pertaining to an observation would be written in asterisks (* analytical comment *), citations would be written in quotation marks (“citation”), and each observation would be accompanied by a time stamp to indicate exactly when a given action or observation happened allowing us to follow the temporal progression of the fieldnotes (see Picture 2). While the time stamps might seem to only constrain the observer further in the field, they indeed turned out to be valuable to the collaborative element of our project. They allowed us to pin-point tendencies temporally in the fieldnotes and compare them across ethnographers to see if the tendencies were in fact patterns. With the Ethno-platform and these common formalities, we now had a framework for our ethnographic work which would ensure an alignment of our notes.

Picture 2. Example of observation in the Ethno-platform

Having these three formalities ensured consistency in our ethnographic data giving us the opportunity to compile fieldnotes i.e., patch together all observations collectively. Of course, the content of each fieldnote is still characterized to some extent by the authoring ethnographer as we have different views and take notice of different things in the field. However, with a firm infrastructural framework in our hands, we establish a common ground for how to note down our observations, and thereby, we have a general format for compiling and storing fieldnotes across a big team of ethnographers. Now that we have established the common structure, in the next installment we will move on to how we ensured that our fieldnotes not only align in format but also in content.

Notes

[1] At SODAS, we are currently in the process of developing our own web-based application with similar, but more user-friendly features. However, for the pilot version in 2021, the software was provided by Survey Exact and ran through an internet browser.

[2] General Data Protection Regulation (GDPR) is a regulation to privacy law in the European Union (EU) that protects personal data of EU citizens.

DISTRACT

The authors if this blogpost series are Sofie Læbo Astrupgaard — PhD fellow in Social Data Science at the University of Copenhagen, Clara Rosa Sandbye — PhD fellow at the Department of Anthropology at Aarhus University, and Emilie Gregersen — MSc student in Social Data Science at the University of Copenhagen. The trio has been working as a part of the interdisciplinary research project DISTRACT, studying the dynamics of issue attention at a political festival.

Source

Interdisciplinary Data Collection at a Political Festival 

Under a gleaming tent canvas by the shores of a small Danish island, a panel of speakers discuss the current housing facilities for elders in Denmark. They interchangeably laugh, argue gravely, gesticulate, look at each other, look at the debate facilitator, look at the audience. Some spectators follow the discussion with great interest; some react to the country tunes that flow from the direction of the main stage; others look at their phones; one fumbles for mints in her handbag; and yet, one has surrendered to the persuasive drowsiness of his hungover body, eyes closed and head resting on his shoulder… And there you are, an ethnographer in the midst of it all. What do you note down? The presence of six colleagues jotting down notes at other stages around the festival site reminds you that the words and numbers you write have a purpose. And a very specific one. They must be understood and interpreted by other researchers as well as machines…

In June 2021, we ventured to the Danish Island of Bornholm to conduct a week-long interdisciplinary case-study of the political festival The People’s Meeting. This is an annual three-day festival involving public and private stakeholders organizing events for common civilians and decision-makers to participate in and discuss current issues. We were a total of 10 researchers all part of the SODAS[1]–based and ERC-funded research project, DISTRACT, which sets out to study political attention economy in Denmark, and The People’s Meeting was the perfect setup to do so.

A key goal of DISTRACT is bringing together theories and methods from different social science disciplines and data science. We were a handful of anthropologists in an otherwise interdisciplinary team who wanted to collect data about attention-related behavior among audiences at events around the festival site. This was complicated as the ethnographic data had to meet the interdisciplinary aims of the study and be integrable with other types of data such as register data, survey data, and data from the festival website and social media. On top of that, we wanted to utilize the programming muscles of some of our team members to computationally process the collected ethnographic data. While such aspirations excited the team, they also came with a one-million-dollar question to be answered: “How does one produce ethnographic data which is both comparable across researchers, compatible with other data types collected in the project, while also holding potential to be computationally processed?”  We were in dire need of a much more structured approach to the classical ethnographic data collection than what we previously had embarked upon if we wanted to answer this question.

An Untraditional Approach

As anthropologically trained junior scholars we have been taught to gather thick descriptions of what we encounter as advised by Clifford Geertz. While we value this approach to the field, our aims as an interdisciplinary team called for a diversion from this notion of “thick” data. Departing from traditional approaches at first seemed like making cracks in our own disciplinary backbone. However, the setting of a chaotic festival site in addition to our interdisciplinary aspirations called for a more preplanned and structured approach if we wanted to collect ethnographic data which would fulfill our ambitions. Of course, this diversion has its costs as collaborative and formalized fieldwork requires the ethnographer to constrain themselves to a common focus dismissing potentially important situations in the field. We’ll delve much deeper into this in a discussion of pros and cons towards the end of the series. For now, we’ll just emphasize that with this piece we don’t wish to suggest a reformation of thick, ethnographic data as we know it and say definitively “goodbye”. Instead, we say “See you later, thick data”, as we merely intend to set aside the classical anthropological approaches temporarily to present a different and more structured take, involving a shift from thick to what we term broad data. By broad data, we mean ethnographic data that fit an interdisciplinary collaborative setting by fulfilling the following qualities (the three Cs): namely, that data can be Compiled in a common format, Compared between researchers, and holds a potential for Computational processing.

Analytical Constructs had to be Considered

Before diving further into our methodology, let us briefly introduce you to what we ventured out to study. Our analytical framework for studying attention flows at the political festival was inspired by the body of micro-sociological theory concerning interaction rituals. In this piece, we won’t dwell so much on the content of these theories, but for the purpose of explaining our methodological approach we’ll provide a brief description: this literature draws on insights from Émile Durkheim and Erving Goffman, and scholars like Randall Collins, Thomas Scheff, and Jonathan Turner are highlighted as main contributors. These theorists formulate general rules for human interaction. The central claim is that if a number of circumstances are at play in an interaction between two or more people, the encounter will result in a common bond of solidarity and flows of emotional energy in the persons who are present. One of the key ingredients for a successful interaction ritual is a common focus of attention. This was our main interest, and methodologically, this meant looking closely at how people interact in micro-situations, e.g., their bodily gestures, and where their visual and auditory attention were oriented. To pick up this type of behavior required a great deal of preplanning and team briefing if everyone was to adopt the same analytical focus when entering the field individually. This will, hopefully, be clear through some of the examples from the field and the methodological considerations that we present in the following sections. But in the meantime, we hope you’ve got your appetite awakened as we’ve only just begun to unfold our journey.

In the coming posts, we will introduce you to our so-called Ethno-platform, a self-developed digital platform for writing field observations. You will dabble further with the notion of broad data as we take you through a discussion of each of the three Cs. Lastly, we will discuss what is lost and what is gained when one diverts from more traditional ethnographic data collection methods. Stay tuned.

Notes

[1] SODAS is the Copenhagen Center for Social Data Science at the University of Copenhagen

DISTRACT

The authors if this blogpost series are Sofie Læbo Astrupgaard — PhD fellow in Social Data Science at the University of Copenhagen, Clara Rosa Sandbye — PhD fellow at the Department of Anthropology at Aarhus University, and Emilie Gregersen — MSc student in Social Data Science at the University of Copenhagen. The trio has been working as a part of the interdisciplinary research project DISTRACT, studying the dynamics of issue attention at a political festival.

Source

Sofie Læbo Astrupgaard is a PhD fellow in Social Data Science at the University of Copenhagen, and she holds a BSc in Anthropology. Sofie’s research focuses on hybrid workplaces, and she likes to explore how data from her ethnographic fieldwork can be combined with large scale unstructured data in meaningful ways in her research.

Clara Rosa Sandbye is a PhD fellow at the Department of Anthropology at Aarhus University, where she does research within the field of restorative justice, concerning issues of criminalization, violence, and morality. She likes exploring the possibilities of collective ethnography, interdisciplinary, and mixed-methods research.

Emilie Gregersen is an MSc student in Social Data Science at the University of Copenhagen and holds a BSc in Anthropology. Her interests include experimenting with traditional ethnographic methods in combination with computational tools and methods from other disciplines, and she hopes one day to become her own version of a computational anthropologist.

See You Later, Thick Data

This blogpost is the introduction to the methodological series “See You Later, Thick Data – How we experimented with doing collaborative fieldwork as part of an interdisciplinary research project”. Through five blogposts, we – a group of anthropologically trained junior scholars – discuss some of the opportunities and challenges we faced when collecting data in an interdisciplinary case study of the Danish political festival The People’s Meeting. We took on a somewhat different approach to ethnographic fieldwork, and in this series, we share our experiences with a highly systematic and collaborative way of collecting ethnographic data that is integrable with different data types. In this introduction, we present the context of our study, including a description of The People’s Meeting and the emerging field of machine anthropology.

Thrown into a world of social data science

Our interest in the field of social data science started back in 2019 as an extra-curricular project, taking up our spare time while writing up each of our bachelor’s theses in anthropology at the University of Copenhagen. We had volunteered as “test subjects” at the Copenhagen Center for Social Data Science (SODAS) and we wondered what it would take for a group of anthropology students to learn how to code. This was how we were first introduced to the curious combination of anthropology and computational methods.

A group of anthropologists takes a crash course in programming. It almost sounds like the first phrase of a bad joke: How many anthropologists does it take to make a function in the programming language Python? It took some work and some getting used to, that’s for sure. Shortly after the course, we came to work at SODAS for several years. The groundwork for this blogpost series is a research project we took part in while working at SODAS.

In brief terms, SODAS is an interdisciplinary research and education center at the Faculty for Social Sciences, UCPH. The center houses researchers from disciplines as diverse as anthropology, economics, sociology, political science, psychology, and data science. The aim is to expand the scientific toolbox by introducing research methods from data science to social science research. The digitalization of societies creates immense quantities of digital data that offer important insights into social life today – and SODAS has deemed it its mission to utilize this data in social scientific research.

Being thrown into the world of social data science taught us important lessons about anthropology. The clash with other methodologies, epistemologies, and data types forced us to consider fundamental questions: What are the qualities of anthropological approaches and ethnographic methods? What aspects of social life can they illuminate, and what aspects call for other methods? And maybe most important; how can these different methods and diverse data types be combined in a meaningful way?

Machine anthropology – Bringing together anthropology and data science

As junior anthropology scholars with a fondness for coding, we are indeed inspired to by machine anthropology. The term has been coined by our supervisor Morten Axel Pedersen to capture ongoing attempts by interdisciplinary research teams at SODAS and elsewhere to explore what an integration between anthropology and data science might look like[1]. Drawing on but also going beyond recent attempts to mix thick ethnographic data with thin big data (e.g. Isfeldt et al. 2019), machine anthropology aspires to unsettle established disciplinary, methodological and epistemological boundaries by using computational methods for augmenting and automatizing the collection, processing and analysis of ethnographic data, and vice versa. The approach we present in the blogpost series can be considered machine anthropology as we attempt to gather ethnographic data that is integrable with quantitative data types, and this implies that the data can be computationally processed.

The research project – DISTRACT and DISTRACT Politics

The study is part of two research projects: DISTRACT: The Political Economy of Attention in Digitized Denmark and ”Ethnographic Text as Data”[2]. DISTRACT brings together diverse social science and data science methods to explore the mental, social and material techniques by which attention is captured, retained, and distracted in the world’s most digitized country, Denmark. One subproject, DISTRACT Politics, combines qualitative (e.g. fieldwork observations) and quantitative data (e.g. social media data) to map and mine the dynamics of political events (Meinert & Kapferer 2014), and to contribute to sociological work on how “issue attention” flows between politicians, media and publics across digital and non-digital media (e.g. Barbera et al 2019). Ethnographic Text as Data seeks to experiment with the use of computational methods for the collection, processing, and analysis of ethnographic data; to develop a theoretical framework for a future computational anthropology, and to contribute to and expand the quali-quantitative toolbox for the social scientific study of political processes and events. The research project our blogpost series is based on falls under the Political Attention subproject while our methodological framework is developed with Ethnographic Text as Data in mind.

The People’s Meeting as research case

The People’s Meeting is an annual political festival, which takes place in the old fishing village of Allinge on the Danish Island of Bornholm in the Baltic Sea. The festival lasts for four days and was established with the official goal of bringing together citizens and decision-makers and facilitating a democratic dialogue. Within this framework, public and private stakeholders organize events such as debates and speeches. Since the festival was launched in 2011, it has grown bigger each year. In 2019, the festival peaked by attracting more than 114.000 visitors. However, due to covid-19 regulations, the festival was cancelled in 2020, and in 2021, it was considerably downscaled to around 8.000 visitors and 450 events.

The issues debated range from human rights, climate issues, working conditions, public health, and much more. Some events are TV-transmitted and live-streamed, and it’s not unusual that politicians use this occasion to announce new and often substantial political messages. In the past years, the People’s Meeting has been more present in the online sphere as well as stakeholders promote their activities at the festival, while politicians and activists continue discussions raised during debates on social media.

In many ways, The People’s Meeting resembles a microcosm of the political landscape in Denmark. The festival attracts the most influential political actors and stakeholders, as well as the media and smaller, more locally anchored organizations. It also attracts thousands of members of general public, ranging from well-to-do retirees to high-school and university students attracted by the large amounts of free beer and snacks handed out by various stakeholders

The People’s Meeting, then, is a perfect setting to study political attention in Denmark. Indeed, as one of our scholarly collaborators, Lasse Liebst points out, festivals ”offer a natural laboratory” (2019: 30) for systematic empirical social science investigation. In addition, it represents an ideal site for interdisciplinary collaboration and machine anthropological experimentation. So, in collaboration with the non-profit organization behind The People’s Meeting, qualitative and quantitative data pertaining to micro-sociological processes and political attention dynamics pertaining to this “natural laboratory” was what we sought to collect in June 2021.

A DISTRACT expedition to the People’s Meeting

We were a team of 10 DISTRACT researchers, who travelled to Bornholm in June 2021. Our primary goal was to test fundamental social theories about so-called interaction rituals and attention dynamics (we will expand on this theory later). A second objective was to experiment with methods for collecting, processing and integrating radically different kinds of social data ranging from ethnographic fieldnotes to sensor data and social media data. Our team was composed of both senior and junior scholars. Half of us had a background in anthropology, and disciplines such as sociology, economy, political science, and social data science were also represented among team members.

All together, we collected a large pool of different data during the festival. While our ethnographic methods such as observations and interviews are the focus of this blogpost series, as alluded to, our team also collected quantitative and digital data, including sensor data, weather data, social media data, as well as register data and survey data we obtained from the organizers of The People’s Meeting. With this blogpost series, we share our experiences with producing ethnographic data that is scalable and integrable with other data types. This involves an online platform for writing fieldnotes and a highly structured methodological approach.

What is the coming blogpost series about?

Now that we have given you an introduction of both our own journey into the field of social data science as well as the context of our project, let us give you a brief outline of what you can expect from the upcoming methodological blogpost series. It will consist of five posts in total: In the first post, we elaborate on why we chose to diverge from classical anthropological approaches to data collection and instead gather what we term “broad” ethnographic data. By this, we mean data that fits an interdisciplinary, collaborative setting and fulfils the three Cs: Namely, that broad data can be Compiled in a common format, Compared between researchers, and holds a potential for Computational processing. In the second post, we describe how we compiled (1st C) ethnographic fieldnotes across our group by using the self-developed “Ethno-platform”, an online tool for writing and archiving fieldnotes. In the third post, we present a systematic approach to data collection that involves self-developed observation schemes and seating charts. These allowed us to align and compare (2nd C) ethnographic data across ethnographers. In the fourth post, we move on to describe how we retrieved structured data from the Ethno-platform that could be merged with data from the schemes and charts. This enabled us to computationally process (3rd C) our data to explore patterns, and it also allowed us to combine this broad ethnographic data with other data types. In the fifth and final post, we discuss the trade-offs when collecting broad instead of thick data, and we argue that for interdisciplinary collaborations, broad data can be preferable.

The approach to ethnographic data collection we present in this series might challenge anthropologically trained readers, just as we ourselves were challenged along the way. What we call broad ethnography does not offer thick holistic descriptions of people, places, and situations. Neither does its methodology offer much flexibility or deep immersion into a field. For some readers our contribution might even seem slightly blasphemous. However, before rejecting it as so, bear in mind that what we present is not a suggestion to reform anthropological fieldwork. Rather, collecting broad data is suited for short-term, collaborative ethnographic data collection in interdisciplinary research, and for this purpose, we believe it holds a great deal of potential. We hope that our blogpost series will engage our readers and we are looking forward to discussing the trade-offs of broadening thick ethnographic data as we know it.

Bibliographic references

Barberá, P., Casas, A., Nagler, J., Egan, P., Bonneau, R., Jost, J., & Tucker, J. (2019). Who Leads? Who Follows? Measuring Issue Attention and Agenda Setting by Legislators and the Mass Public Using Social Media Data. American Political Science Review, 113(4), 883-901.

Breslin, S., A. Blok, T. Enggaard, T. Gårdhus, and Pedersen, M. A. (2022). “Affective Publics”
Performing Trust on Danish Twitter during the COVID-19 Lockdown. Current Anthropology 63(2).

Isfeldt, A. S., Enggaard, T. R., Blok, A., & Pedersen, M. A. (2022). Grøn Genstart: A quali-quantitative micro-history of a political idea in real-time. Big Data & Society, 9(1).

Liebst, L. S. (2019). Exploring the sources of collective effervescence: A multilevel study. Sociological Science, 6, 27-42.

Meinert, L., & Kapferer, B. (Eds.). (2015). In the event: Toward an anthropology of generic moments. Berghahn Books.

Sekara, V., Stopczynski, A., & Lehmann, S. (2016). Fundamental structures of dynamic social networks. Proceedings of the National Academy of Sciences, 113(36), 9977-9982.

Notes

[1] SODAS is the Copenhagen Center for Social Data Science at the University of Copenhagen

[2]DISTRACT is funded by the Advanced Grant project 834540 from the European Research Council. Text as Data is funded by the Data + Program at the University of Copenhagen.

DISTRACT

The authors if this blogpost series are Sofie Læbo Astrupgaard — PhD fellow in Social Data Science at the University of Copenhagen, Clara Rosa Sandbye — PhD fellow at the Department of Anthropology at Aarhus University, and Emilie Gregersen — MSc student in Social Data Science at the University of Copenhagen. The trio has been working as a part of the interdisciplinary research project DISTRACT, studying the dynamics of issue attention at a political festival.

Source