16  Thematic analysis (Jo-Ju Kao)

16.1 Folien

Die Folien sind auf Blackboard🔒 verfügbar.

16.2 Transkript

WarnungHinweise zum automatisiert erstellten Transkript

Das folgende Transkript wurde auf Basis der Aufzeichnung der Vorlesung erstellt. Die vollständigen Aufzeichnungen inklusive der Bildschirminhalte sind in Blackboard🔒 verfügbar. Die Tonspur wurde zuerst mit Hilfe der Werkzeuge des Oral-History.Digital Projekts wörtlich transkribiert. Die wörtliche Transkription wurde in Kombination mit den Vorlesungsfolien mithilfe von Sprachmodellen (v. a. Claude Sonnet 4.5 und GPT 5.2) zu einem übersichtlichen Transkript zusammengefasst. Im Anschluss wurde das Transkript von einer studentischen Hilfskraft überprüft, geglättet und ggf. angepasst. In diesem Prozess kann es an verschiedenen Stellen zu Fehlern kommen. Im Zweifel gilt das gesprochene Wort, und auch beim Vortrag mache ich Fehler.

Ich stelle das Transkript hier als experimentelles, ergänzendes Material zur Dokumentation der Vorlesung zur Verfügung. Noch bin ich mir unsicher, ob es eine sinnvolle Ergänzung ist und behalte mir vor, es weiter zu bearbeiten oder zu löschen.

Thematic Analysis

This lecture presents thematic analysis as a flexible qualitative method for identifying, analyzing, and interpreting patterns of meaning in a dataset.

What Thematic Analysis Is

The lecturer introduces thematic analysis as a widely used but poorly demarcated qualitative method. It is described as a method for identifying, analyzing, and interpreting patterns of meaning, also called themes, within a dataset. In practice, it means reading and rereading the material, labelling it, categorizing it, synthesizing it, and interpreting it.

The lecture also explains that thematic analysis is often a first useful qualitative method because it teaches core analytical skills that can be transferred to other qualitative approaches. In very simple terms, it helps researchers sort data into meaningful groups.

• Code: a short label that summarizes something important in the data.
• Theme: a more abstract pattern that captures something important in relation to the research question.
• Iterative process: the analysis moves back and forth between steps rather than following a straight line.

Research Questions

The lecture makes clear that thematic analysis can be used for a wide range of research questions. A research project always has an overall question, but that question may become more focused during analysis. The lecturer emphasizes that research questions do not have to be fixed from the start; they can evolve as the analysis develops.

Three kinds of questions are highlighted: questions about what themes exist, questions about how something is communicated, and questions focused on a specific concept of interest. This makes thematic analysis especially suitable for broad, exploratory projects.

From Words to Codes

The first analytical step is moving from words to codes. A code is a higher level of abstraction than the more concrete information in the material. This means the researcher begins to interpret and generalize instead of staying too close to the literal wording.

The lecturer stresses that this process always involves a trade-off: some specific detail is lost, but the analyst gains the ability to detect patterns and connections. Good codes should be clear, distinct, non-overlapping, and each should capture a unique idea.

The lecture contained multiple exercises to illustrate the different steps of thematic analysis.

Exercise 1: Sustainable Households

The first exercise asks students to code a paragraph about a household trying to live more sustainably. The text mentions many waste bins, different collection days, and the need to bring food waste to a communal composting site. It also describes a positive outcome, because the daughter becomes involved in the community garden, learns new things, makes friends, and the household receives fresh vegetables and herbs in return.

The lecturer uses this example to show that coding depends on the research question. Possible codes might focus on waste separation, practical inconvenience, community involvement, or environmental reward. The lecturer also notes that some codes, such as food, paper, and glass, could already be grouped into a broader category like waste.

Exercise 2: Coding Style Variation

The second exercise uses a short excerpt about a boy named Barry who goes through a difficult school period. He is described as a people-pleaser, admired by teachers, then harmed by the influence of two boys who put him down, which he internalizes, and he later loses his social network. Students are asked both to identify a possible research question and to assign codes.

The lecturer uses this passage to show different coding styles. Descriptive codes might be student-teacher relationship or bad role models. Direct quotation codes might stay close to the wording, such as tough time or people pleaser. Process coding might capture movement and change, such as internalizing rejection or losing one’s network. The point is that no single style is always best, and combining styles can give a fuller picture.

From Codes to Themes

After coding, the next step is moving from codes to themes. Codes are closer to the data and simpler, while themes are more interpretive, more abstract, and more closely linked to the research question. To build themes, the researcher compares codes, groups them, and analyzes how they relate to one another.

The lecture suggests asking questions such as what the code is about, what it is getting at, which other codes it relates to, what is common across a group of codes, and how the codes connect to theory, literature, and the research question. The lecture also notes that not every code has to fit neatly into one theme, because thematic analysis requires judgment and consistency rather than fixed counting rules.

Exercise 3: Vegan Students

The third exercise asks students to group already coded interview data from a study about students talking about becoming vegan. The lecturer invites students to think about how codes might be clustered into broader themes. The example themes mentioned include incentives to continue eating meat and obstacles to becoming vegan.

The lecturer explains that it is normal to find some codes difficult to place. Not all codes will be used in the final analysis, because the research question determines which aspects are most relevant. This is why thematic analysis always requires selectivity.

Thematic Map Example

The lecture also introduces thematic maps as a way of organizing codes into themes. One example concerns a study of women talking about their vagina. In the first stage, everything is still connected. In the second stage, the analyst begins to separate and refine the material. In the final stage, the analysis is reduced to two main themes: positive talk and negative talk.

This example shows that theme development is a dynamic process. The analyst does not simply list every code, but gradually groups, revises, and sometimes drops parts of the material until a coherent structure emerges.

The Full Process

The lecture summarizes the main phases of thematic analysis as familiarizing yourself with the data, generating initial codes, searching for themes, reviewing themes, defining and naming them, and writing the report. These phases are not strictly linear. Instead, the researcher repeatedly moves back and forth between them, especially when themes do not fit well and codes need revision.

The lecturer describes this as a pattern-finding process that happens through reading and rereading. Reworking the analysis is not a mistake or wasted time; it is a normal part of the method.

Examples in Communication Research

The first example discussed is a study on ambient digital racism and racial narratives on Twitter. The researchers collected tweets using hashtags such as WhiteLivesMatter, BlueLivesMatter, and AllLivesMatter, and analyzed 203 tweets. Their aim was to identify and analyze racist discourses on Twitter in the context of George Floyd’s death. The results included themes such as oppressors reverse racism and a counter-narrative based on the social criminalization of BLM.

The second example concerns aspirational content creators’ narratives about YouTube’s algorithm on Reddit. The researchers studied 144 Reddit posts and used inductive coding. They then refined broader themes through meetings and collaborative discussion, showing how thematic analysis can be done transparently and iteratively. The lecture also addresses a question about whether one code can belong to more than one theme, and the answer is yes, as long as this is defined and justified in the method section.

The third example looks at how Black communities used social media as a space for healing during COVID-19. The researchers analyzed 2,000 Instagram comments from two online events designed to support Black audiences. They identified 13 themes but discussed only the eight that directly answered the research question. The lecturer presents this as a good example of being analytically selective instead of merely descriptive.

Why Use It

The lecture explains that thematic analysis is especially useful in communication research because it is often inductive and data-driven. It is well suited to exploratory research and to emerging or underdeveloped topics. Its flexibility is also a major strength, because it is not tied to one specific theory in the way framing analysis or discourse analysis often are.

Another advantage is that it is relatively easy and quick to learn, even for researchers with little qualitative experience. It can also summarize key features of data and provide a thick description. The lecturer specifically suggests that it can be a very sensible method for bachelor’s and master’s projects, especially when the aim is to make sense of new communication phenomena.

Why Not Use It

The lecture also gives several reasons not to use thematic analysis in some cases. If a topic is already well developed and there is abundant existing literature, another method may be more appropriate. For example, if predefined categories already exist, quantitative content analysis may be a better fit for grouping data.

The lecturer also notes that thematic analysis lacks strong methodological guidance compared with some other methods. It does not allow strong claims about language use in the way framing or discourse analysis can. Its flexibility can also become a problem if it leads to inconsistency or lack of coherence.

Doing It Well

To do thematic analysis well, the lecturer emphasizes close reading, repeated reading, and systematic coding. Researchers should reflect on what they are doing, be able to explain and justify their decisions, and keep revising codes and themes as needed. It is also helpful to write notes before, during, and after coding.

The lecture also discusses self-assessment questions. The first group concerns relevance: what are you coding for, what do you include in your thematic map, and does it fit the research question? The second group concerns labelling: are the names of your codes and themes precise and useful, and are you using concepts appropriately?

The lecturer further warns against describing themes as simply “emerging” from the data. This language hides the active role of the researcher in identifying patterns, selecting what matters, and reporting the findings. Instead, the analysis should be understood as an interpretive process in which the researcher makes choices and should acknowledge them as such.

Trustworthiness

A key quality concept in thematic analysis is trustworthiness. It is presented as parallel to reliability and validity in quantitative research and includes credibility, transferability, dependability, and confirmability. The central principle is transparency.

To strengthen trustworthiness, the lecturer recommends keeping audit trails that document decisions, explanations, and justifications. In other words, the analysis should be clear enough that readers can follow how the themes were produced.

Final Summary

The lecture concludes that thematic analysis tries to grapple with, simplify, and articulate a messy reality. The goal is to be systematic, consistent, and transparent while still producing a rich interpretation that answers the research question. The method is therefore presented as practical, flexible, and especially useful for studying new communication phenomena in a structured way.