Analyzing Ethnographic Data

Analyzing Ethnographic Data

April Burns


In this chapter, I will outline some principles of data analysis that will guide you in analyzing, interpreting, and writing up your ethnographic research projects. Skilled ethnographers are not unlike creative professionals, such as designers or chefs, in that their work products come together as a result of creativity, sustained effort and giving up some control over the final product. While cooks tend to follow recipes, chefs follow culinary principles, foundational truths about the types of flavors that work well together, and the practices that create delicious dishes. When you follow a recipe, you are limited to a specific dish, you follow someone else’s plan based on their priorities and tastes. This can be useful as you learn the fundamentals of cooking. However, it is more liberating and useful to become fluent in the ideas that make cooking a range of dishes (or analyzing a range of social settings) possible. Similarly, these principles and practices of data analysis will open the doors to the meaningful interpretation of all sorts of ethnographic data. Though there may be some recipe-like directions in this chapter, my goal is to instead emphasize the most critical principles of data analysis.

What is ‘analysis’ and why do we do it?

You’ve collected data and begun to put together many of the pieces of a social cultural puzzle, but the pieces by themselves don’t yet tell a story. Data analysis is the stage in which you work to put the pieces together to present a picture of the social context that you set out to study. It is the process by which we make sense of all the data that we have collected over the course of a project (Bailey, 2018, p. 159) specifically in order to make some assertions about a particular social and cultural space. These assertions should be based on the various data that you have collected including the observations you conducted, the experiences and the interactions you engaged in, and the fieldnotes you wrote documenting such observations, experiences and interactions. Data also include the interviews you conducted and transcribed, the artifacts and official documents you have collected, any photos you have taken and any maps you may have drawn. Brewer (2000) defines analysis as “the process of bringing order to the data, organizing what is there into patterns, categories and descriptive units, and looking for relationships between them” (p. 105). This analysis stage is made possible by our previous data organizing (sometimes called data reduction) and engaging in the coding process, both of which facilitate an intimate familiarity with our data. From there, data analysis should move toward ‘interpretation,’ which “involves attaching meaning and significance to the analysis, explaining the patterns, categories and relationships” (Brewer, 2000, p. 105).

Data analysis, like much creative work, is often improvisational and therefore, a “messy” process (Murchison, 2010, p. 181). A successful outcome will likely require you to ‘lean into’ what can feel like a chaotic endeavor. Accepting such ambiguities and uncertainties is part of the process of becoming an ethnographer, indeed a researcher of any sort. While coding and analysis can be a “messy” and thus challenging process (Bailey, 2018, p. 160), keep in mind that we do versions of this a lot in our lives. Consider for example, how we go about cleaning a messy room, desk or closet. We first organize the contents of the room, desk or closet into categories that make sense to us, our purposes, and our lives (e.g., Clean clothes, books). This is like the coding process and the contents of the space are the separate pieces of data we have collected. Think about undertaking such a process for someone else, tidying their closet for them. After organizing/categorizing their stuff, you will likely learn a lot about them and additional questions will have emerged. Why do they have so many pairs of shin guards? What occasions do they wear their formal clothes to? Why are three boxes of letters important enough for them to have kept and organized by month and year? The answers to such hypothetical questions will tell you even more about the lives and experiences of the closet owners. The process by which you become familiar enough with the closet contents to ask questions and work to answer them, and then craft an explanation of the meaning to the closet owner, constitutes the analysis stage.

Keep in Mind:

Your unique and informed perspective is valuable and could potentially reveal aspects of a social context that previous researchers hadn’t considered. As the creative and intellectual force leading this process, you are free to innovate in how you apply these principles to your research and assignments, when and where it makes good sense to do so – as long as you document and explain your reasoning behind such revisions.

Stages of Analysis:

Coding è analysis begins thru evolving coding è coding moves to identification of key themes è analysis develops thru memo-ing è analysis continues toward interpretation è leading to assertions/theory è written ethnography.

Ethnographers typically move from observing and recording the concrete actions and interactions of people and groups, to organizing this recorded data, to generating ideas conceptualizing and interpreting these practices, to making assertions that relate to larger social and cultural phenomena. This process is described below.

Analysis begins with coding

Ethnographers Scott-Jones & Watt (2010) have described ethnographic analysis as a two-stage process first involving the organizing and ordering of your data, and then the analysis proper (Scott-Jones & Watt, 2010, p. 159). In their framing, coding heralds the transition between these two stages. Similarly, O’Reilly (2012) distinguishes between the “writing down” (i.e., collecting/recording of data) and “writing up” (i.e., presentation) phases of ethnographic research, locating analysis as the stage in between in between these two phases, although aspects of analysis are present in both stages.

Despite some variation in the number of processes or stages that data analysis is broken down into, the analysis of data generally begins with the “coding cycle” that you read about in the chapter by Torres-Rivera (2019) [add link]. Torres-Rivera describes the coding process as a “repeating cycle of three phases: the coding phase, writing memos phase, and reviewing/revising/refining codes phase” (pg #). In this text, we decouple coding from analysis but really, coding is the first stage of analysis, and the line between coding and analysis is somewhat subjective. While I won’t be rehashing the material in the coding chapter, there are some important aspects of coding, particularly the iterative as well as inductive aspects outlined by Torres-Rivera, that are worth repeating here as they are also aspects of data analysis overall.

O’Reilly (2012) similarly describes ethnographic research as “iterative-inductive,” that is, a “practice of doing research, informed by a sophisticated inductivism, in which data collection, analysis and writing up are not discrete phases, but inextricably linked” (p. 180). The ‘iterative’ aspect means that analysis is ongoing, resembling a spiral process rather than a circular or linear process (O’Reilly, 2012, p. 181), while the ‘inductive’ aspect means that researchers work to make general assertions from the specifics of their data. This combined “iterative-inductive” concept highlights that a researcher’s analytic choices are ongoing, informed by what they learn as they progress through the stages of the research process, from the early decision to pursue a particular research question and choice of a research design, to the identification of certain field sites and coding scheme used to organize data.

Thematic Analysis:

Codes often begin as descriptive markers of things people say (e.g., in vivo coding) or do (e.g., process coding) (Bailey, 2018, p. 161; Brewer, 2000, p. 110). To transform our coding efforts into analysis, requires that we refine our coded categories, typically through multiple rounds of the coding cycle (Madden, 2017, p. 142), leading to the identification of key themes within your data (Brewer, 2000, p. 109; Madden, 2017; Saldaña, 2011, p. 108; Scott-Jones & Watt, 2010, p. 162). While codes can develop into themes, you can also directly code your data for themes, as thematic analysis is a common method of data analysis for ethnographers and other qualitative field researchers (Bailey, 2018). Themes differ from codes in that themes are generally longer phrases that attempt to convey the manifest (or apparent) meaning of data as well as the latent (or underlying) meanings of data (Saldaña, 2011, p. 108).

According to Madden (2017), a theme “could be a large sociological category, a group behavior, an individual behavior, an aspect of the physical setting or an observation of a mood or feeling (..)” (143). Bailey (2018) breaks down the broad category of themes into two main types: “topical” themes and “overarching” themes.  This distinction is useful because it leads us into an important task of analysis, which is conceptualization, which entails moving codes from particular and concrete actions, to a more abstract level of concepts and ideas (Bailey, 2018, p. 165). Conceptualization allows us to later build connections to theory, which in EOW, are typically theories of work and economy. Overarching themes require us to zoom out from the specifics of what people say and do (as captured in codes and topical themes), to highlight those ideas or concepts that these words and actions suggest or represent in overarching themes.

Themes exist in every film you’ve ever seen and in every book you’ve ever read. You enjoy specific scenes, characters, plot twists and dialogue, but those specific things are not what the movie is about. When describing a film, we do so in terms of what it means and the messages it conveys: a ‘coming of age story,’ a story of ‘love and loss,’ a story of ‘redemption,’ of ‘moral reckoning,’ of ‘cultural annihilation.’ These broad concepts are versions of overarching themes. We can assign overarching themes to our data as well, though it is rarely a straightforward process and often more than one theme could be assigned. To uncover the overarching themes in a specific piece of data, such as an interview or set of field notes, imagine tweeting or texting its meaning to your social network. How would you summarize the underlying message or meaning of that form of data, in a just sentence or two? To do that, you would have to follow the process below.

How we identify themes:

While “there is no universal template” for the process of identifying themes (Madden, 2017, p. 145), there are a number of commonly accepted strategies for identifying key themes in your data, including by taking note of how often certain codes or themes show up in the various forms of your data (Bailey, 2018, p. 188; Murchison, 2010, p. 116). Those that show up more frequently warrant consideration because the more that something is observed or spoken about, the greater its potential relevance and impact on the lives you are studying. More than two occurrences in the data also presents a pattern open for interpretation, potentially giving us more information about our research question. This does not mean that every phenomenon that happens more than once represents a pattern, or a useful pattern relevant to our research question – a reoccurrence may simply be a coincidence or due to random chance. Still, frequency is an important flag for potential themes.

A fieldwork activity that I assign in my Ethnographies of Work course illustrates this process. Students briefly interviewed seasonal workers at a large holiday market. After collecting student responses to this assignment over the course of a few semesters, I identified the following recurring themes.

“One thing you learned about seasonal workers at a holiday market?”

Seasonal workers: work long hours and most days

Seasonal work is: fast paced & high volume

Seasonal work is: Short-term/temporary

Seasonal workers: enjoy interacting with diverse & international customers

Seasonal workers: enjoy the bustling park environment

Seasonal workers: enjoy their jobs and work

Seasonal work: Is often a ‘side’ job rather than a main/only source of income

Seasonal workers: sometimes relied on personal contacts to get the job

Organizing your themes

Because these themes remain at the material level of what workers said and/or do, these are considered “topical” themes representing the apparent meaning, rather than representing overarching (i.e., more conceptual) themes (Bailey, 2018) addressing the latent meaning (Saldaña, 2011). To move topical themes to the conceptual level of overarching themes, we will need to interpret the underlying meaning that these themes embody. There is no easy recipe for such conceptualizing. However, at this point, you may notice that some coded categories cluster together because they are related in some way.

Revised themes: “One thing you learned about seasonal workers at a holiday market?”


  • Seasonal workers: work long hours and most days
  • Seasonal work is: fast paced & high volume
  • Seasonal work is: Short-term/temporary


  • Seasonal workers: enjoy interacting with diverse & international customers
  • Seasonal workers: enjoy the bustling park environment
  • Seasonal workers: enjoy their jobs and work


  • Seasonal work: Is often a ‘side’ job rather than a main income source


  • Seasonal workers: sometimes relied on personal contacts to get the job


Intentionally organizing related categories into an outline format of superordinate and subordinate categories, or into hierarchies of significance based on the relative importance of the concept, is a useful way to more finely understand specific phenomena, behavior, or experiences (Saldaña, 2011, p. 108). Creating a “hierarchy of significance” (Scott-Jones & Watt, 2010, p. 162) will allow us to consider how such coded categories are related and how they possibly interact (Saldaña, 2011, p. 98). I have thus grouped the topical themes (above) into four categories, with the first two being primary categories based on their prominence in the data, and that make the most immediate sense to me: job qualities and feelings about the job. The last two themes in my list, addressing how some workers got their job and the role that the job plays in their overall employment, appeared less frequently than the first two sets of themes, and were thus placed lower down the list. But I also wasn’t yet sure how to organize these last two in relation to the other theme categories, so they remained individual topical themes for now.

This basic interpretive framework applied to the student data on seasonal workers allows us to consider each group of themes as its own thematic category, and to further organize these categories in a way that makes clearer conceptual sense and helps us to understand this specific context of work. Looking at the job qualities themes, I noticed that 2 out of 3 themes center around the intense demands of this type of work, so I renamed (below) this category “demanding work.” The short-term aspect of this type of seasonal employment doesn’t seem to fit as well in this newly named category, which required me to re-read the data, write notes, and revise my framework. Looking at the second set of themes lead me to rename this category, “job satisfaction.” Remember that I wasn’t sure how to arrange the last two themes in the list, so I labeled them as individual and separate themes: “networking” and “side job.” However, after reevaluation, the short-term/temporary theme that I first put under “demanding work,” seems to fit better in the “side job” category, given that it also references a secondary and contingent aspect of this type of employment.


Seasonal workers: work long hours

Seasonal work is: fast paced & high volume



Seasonal workers: enjoy interacting with diverse & international customers

Seasonal workers: enjoy the bustling park environment

Seasonal workers: enjoy their jobs and work



Seasonal work is: Short-term/temporary

Seasonal work is: not typically the main income source



Seasonal workers: some got the job through a personal contact


This revised version makes better conceptual sense and allows me see patterns in the data, and to ask questions about the relationship and interaction between these themes, and the degree to which they exist in tension or harmony. In the example above, there seems to be an apparent contradiction between the demands of the job (which are high) and job satisfaction (also high) that workers themselves did not attribute to high (or low) wages. Such contradictions can flag important concepts and dynamics in action. In other words, why might workers enjoy a job they that they also describe as demanding? Encountering such contradictions within your data offers an interesting analytical opportunity (Murchison, 2010, p. 181), as they require an explanation, thus advancing your analysis to the interpretation level (Madden, 2017, p. 151).

Considering the other key themes specifically in relationship to one another (Saldaña, 2011, p. 92) can help explain this contradiction. Perhaps the fact that this job is short term and a source of extra income (the side job theme) mitigates the demanding aspects of the work because workers understand that the work has a finite and foreseeable end. Another explanation may be the fact that many workers reported landing their job through personal connections, putting the demanding nature of the job in the context of existing relationships that workers likely hope to maintain. Thinking – and writing — about these patterns and relationships also allows me to make some assertions explaining this sociocultural experience.

Additional ways to identify themes:

It’s also important to consider aspects that are absent or not present in your data. While studying what’s not there may seem counter-intuitive, considering this aspect can lead to some important insights. For example, if you were researching the experiences of new and first-time parents, and you never heard any talk about the difficulties adjusting to the new demands, but witnessed what looked like challenging experiences – this may lead you to ask questions about why people haven’t discussed such expected experiences. Or consider the experience of working in a fast paced and high-pressure occupational field, such as the seasonal holiday market, while none of your participants specifically mentioned stress management, might lead you to questions about why such expected issues haven’t surfaced.

Another way that a theme can be identified lies not in how often it comes up in the data, but instead stands out because something you observed people say or do seems to hold particular meaning for the people you are studying. This may include a critical incident, a vivid case, or a strong emotional response observed.

Let’s not forget the importance of your research question, which acts as a north star for researchers, guiding your coding and analysis, and centering the final written product. It may be that the question you are interested in answering doesn’t frequently emerge in your data, so it is important that you intentionally evaluate your data for themes related to your research question. Your research question is a guide but a flexible one, one that you should be open to revising and refining to reflect the insights you are gaining from the analytic process. Your data and analysis may suggest that you haven’t got the right question and you may choose to revise the research question that you started out with (Murchison, 2010, p. 120). Your final ethnographic paper should center around addressing your revised research question.

Analytic Memos: “one of the most useful and powerful sense-making tools at hand” (Miles, et al, 2019, p. 89).

A strong analysis depends on the quality and quantity of the data that you have collected, but it also depends on the level of effort you put into the analytic practice of coding and memo-ing. Analytic memos, described as a, “brief or extended narrative that documents the researcher’s reflections and thinking processes about the data” (Miles, et al, 2019, p. 88), is an integral aspect of data analysis. Writing isn’t just for the final presentation of your work, it is also a means of developing your work. Writing is thinking, and as a “sense-making tool,” memo writing encourages “little conceptual epiphanies” (Miles et al, 2019, p. 89).

Analytic memos are informal and open-ended written explorations of what you are learning from your data and what you still need to learn (Saldana, 2011, p. 98). Consequently, there is no need to worry about grammar, spelling or even full sentences at first as you are the primary audience for these memos. The priority is to document and process the insights that you are gaining. Memos can and should later be revised, extended, combined and edited as they may ultimately make their way into your final public writing.

Memo early and often.

As soon as you begin going into the field and collecting data, you should also begin writing analytic memos (Murchison, 2010). In your ongoing memo writing, it is recommended that you keep a running list of the all the things you have learned about your research setting and question, the things you still want to learn (i.e., emerging questions) and the things that you still need to do (Murchison, 2010). Any of these lists can productively be organized into a hierarchy of significance, from most to least important, leading to increasingly deeper levels of analysis, a more focused research question, and ultimately to a refined understanding of an area of inquiry.

Your early analytic memos will likely be descriptive in nature, which is a fine place to begin your analysis. According to O’Reilly (2012), describing your ethnographic findings “is a crucial phase of the analysis of your data, that often does not reveal itself to you until you start to write things up” (p. 193-194). Writing detailed descriptions of key settings, events, people, and/or exemplary cases are useful and easy entry points into your analysis (Brewer, 2000, p. 111-114; O’Reilly, 2012, p. 193). Jamie Woodcock’s (2017) research in an urban call center in the UK is a good example of the way many ethnographers use detailed descriptions to invite readers into their analysis.

In a chronological narrative, Woodcock (2017) tells the story of his entry into the world of telemarketing, describing his process of getting work at a British call center and detailing what his training was like. He presents three especially difficult customer calls to illustrate the emotionally challenging nature of the call center work. His ‘thick description’ lays the conceptual groundwork for the rest of his ethnography, including his theoretical framing, which includes a discussion of Taylorism/scientific management and emotional labor demands.

Memos should always be dated so you know where in your analytic process each memo is situated. Memos should also be titled and subtitled – this is important because titles (especially the main memo title) summarize the content and/or purpose of the memo, while a more descriptive subtitle narrows your focus to a small area of analysis. Dating and titling will also help you locate and organize your collection of analytic memos later on. A rich collection of memos allows you to look for relationships and patterns across different types of data, such as field notes and interview transcripts as well as within a specific piece of data, such as a long interview or set of field notes. It is especially important to memo after each round of coding, as doing so is an essential part of the iterative coding/analysis process described earlier in the chapter, a process that is hindered if you don’t consistently put pen to paper (or fingers to keyboard) (Saldana, 2011, p. 102). Fully engaging this iterative process requires that you describe and reflect on what you are gleaning from the data and how such insights shape your coding, and in turn, document additional insights gained from your more complex coding.



While descriptive writing is a fine way to start a memo, it shouldn’t be the end point of your memoing practice. Instead, memos are a means of moving you toward the interpretation stage of the analytic process by (Bailey, 2018; Madden, 2017, p. 152). Researchers aim to use their acquired insider perspective, in interaction with their outsider status and information gathering efforts, to offer useful interpretations of the data they’ve collected and constructed. Madden (2017) describes interpretation as “moving from idea to explanation, from data to story, and in many cases from confusion to meaning” (p. 149). Bailey (2018, p. 200), referring to priorities set by Braun and Clark (2006), reminds us that interpretation requires ethnographers to go beyond description to explain the meaning of our identified themes, to point out the beliefs and assumptions that are embedded in identified themes and why it matters, and lastly, to suggest what conditions are associated with the emergence of this theme/concept. This is an opportunity for you to discuss the meaning(s) that you give to your research data, and to explain why you think people behave and think in specific ways (Madden, 2017, p. 151).

Assertion Development

Writing memos allow you to converse with your research literature, data, and the insights you documented and developed in earlier memos, and consequently well positions you to make some assertions about the specific sociocultural setting that you are studying. Making an assertion means declaring or asserting something as truth. You are offering a plausible explanation of the cultural phenomena and/or group experience you have studied that defending this argument with evidence. It is important to acknowledge the many possible, and perhaps competing, perspectives that explain a situation (Brewer, 2000). But as Brewer (2000) points out “some things are less true than others” (p. 122), and some interpretations are more credible and convincing, and thus carry more weight. You’re aim is not to prove your point, but to show how your explanation makes good sense (Saldaña, 2011, p. 125).

Negative Cases/Outliers

Data that are distinct from the majority of responses are called outliers, while data that doesn’t conform to the boundaries of your assertions are known as “negative cases.” These examples have interpretive value and should also be considered in your analysis. Negative cases/outliers can show variation and diversity within a social cultural setting and keep us from essentializing the experience of one person or a group of persons (Brewer, 2000, p. 109; Saldaña, 2001, p. 101).

When you feel stuck

To be successful in almost any endeavor, you must be willing to be uncomfortable – and analyzing data is uncomfortable. Moving into the data analysis stage can be overwhelming, and you may feel like you don’t know where to start or stop. You may wonder whether you have crafted the most relevant or useful codes, worry that you aren’t addressing your research question, or even whether your research question is even any good. You may question your fieldwork and data collection practices (e.g., not enough data, too much data, the quality of the data). These self-questions, while uncomfortable, are actually good signs that you are in the middle of data analysis. The ethnographic researcher’s challenge is to not let such anxieties stop you from doing the analytic work required for a meaningful product. So accept the discomfort, know that it is temporary and is just part of the process that will move you to insights later.

Chapter Summary:

  • Data analysis is the process by which we make sense of all the data that we have collected over the course of a project, allowing the researcher to make some assertions about a particular social and cultural space.
  • Data analysis is an iterative-inductive process. The ‘iterative’ aspect means that analysis is ongoing, while the ‘inductive’ aspect means that researchers work to make general assertions from the specifics of their data.
  • The analysis of data generally begins with the coding process and ends with interpretive assertions, and sometimes theory development, within ethnographic writing.
  • Thematic analysis is a common method of data analysis for ethnographers and other qualitative field researchers.
  • Themes can be separated into “topical” and “overarching” themes or organized into an outline format of superordinate and subordinate categories, or hierarchies of significance, based on the relative importance of the concept.
  • The practice of writing analytic memos after each round of coding, documenting the researcher’s reflections and thinking processes about the data, is an essential part of the data analysis process.


  1. What are some ways to identify key themes in your ethnographic record?
  2. Explain the difference between a topical theme and an overarching theme?
  3. What is the value of writing analytic memos in the data analysis process?
  4. How might you use this memo writing practice in your other coursework? In the workplace or professional setting?
  5. What is the value of considering negative cases and contradictions in the data?
  6. How can a researcher bring their own values and priorities into an ethnographic analysis?

Key Terms:

Analytic Memos


Hierarchies of significance



Key Themes

Overarching theme

Stages of analysis

Thematic analysis




Bailey, C. A. (2018). A Guide to Qualitative Field Research (3rd ed.). United States: SAGE Publications.

Brewer, J. (2000). Ethnography. United Kingdom: McGraw-Hill Companies, Incorporated.

Madden, R. (2017). Being Ethnographic: A Guide to the Theory and Practice of Ethnography. United Kingdom: SAGE Publications.

Miles, M. B., Huberman, A. M., & Saldaña, J. (2019). Qualitative Data Analysis: A Methods Sourcebook (4th ed.). India: Sage Publications.

Murchison, J. (2010). Ethnography Essentials: Designing, Conducting, and Presenting Your Research. Germany: Wiley.

O’Reilly, K. (2012). Ethnographic Methods. United Kingdom: Taylor & Francis.

Saldaña, J. (2011). Fundamentals of Qualitative Research. United Kingdom: Oxford University Press.

Scott-Jones, J., & Watt, S. (2010). Ethnography in Social Science Practice. United Kingdom: Taylor & Francis.

Torres- Rivera, C. (2019). Coding Qualitative Data.

Woodcock, J. (2017). Working the Phones: Control and Resista nce in Call Centres. United Kingdom: Pluto Press.