Chapter 22: Thematic Analysis

 

What is thematic analysis?

Thematic analysis is a common method used in the analysis of qualitative data to identify, analyse and interpret meaning through a systematic process of generating codes (see Chapter 20) that leads to the development of themes.¹ Thematic analysis requires the active engagement of the researcher with the data, in a process of sorting, categorising and interpretation.¹ Thematic analysis is exploratory analysis whereby codes are not predetermined and are data-derived, usually from primary sources of data (e,g, interviews and focus groups). This is in contrast to themes generated through directed or summative content analysis, which is considered confirmatory hypothesis-driven analysis, with predetermined codes typically generated from a hypothesis (see Chapter 21).² There are many forms of thematic analysis. Hence, it is important to treat thematic analysis as one of many methods of analysis, and to justify the approach on the basis of the research question and pragmatic considerations such as resources, time and audience. The three main forms of thematic analysis used in health and social care research, discussed in this chapter, are:

• Applied thematic analysis
• Framework analysis
• Reflexive thematic analysis.

Applied thematic analysis

This involves multiple, inductive analytic techniques designed to identify and examine themes from textual data in a way that is transparent and credible, drawing from a broad range of theoretical and methodological perspectives. It focuses on presenting the stories of participants as accurately and comprehensively as possible. Applied thematic analysis mixes a bit of everything: grounded theory, positivism, interpretivism and phenomenology.²

Applied thematic analysis borrows what we feel are the more useful techniques from each theoretical and methodological camp and adapts them to an applied research context.^2(p16)

Applied thematic analysis involves five elements:

• Text segmentation involves identifying a meaningful segment of text and the boundaries of the segment. Text segmentation is a useful process as a transcript from a 30-minute interview can be many pages long. Hence, segmenting the text provides a manageable section of the data for interrogation of meaning. For example, text segmentation may be a participant’s response to an interview question, a keyword or concept in context, or a complete discourse between participants. The segment of text is more than a short phrase and can be both small and large sections of text. Text segments can also overlap, and a smaller segment may be embedded within a larger segment.³
• Creation of the codebook is a critical element of applied thematic analysis. The codebook is created when the segments of text are systematically coded into categories, types and relationships, and the codes are defined by the observed meaning in the text. The codes and their definitions are descriptive in the beginning, and then evolve into explanatory codes as the researcher examines the commonalities, differences and relationships between the codes. The codebook is an iterative document that the researcher builds and refines as they become more immersed and familiar with the data.³ Table 22.1 outlines the key components of a codebook.³

Table 22.1. Codebook components and an example

Code	Definition	When to use	When not to use	Example
Attitudes or perceptions: falls	Attitudes about falls from health professionals	When a health professional describes their thoughts about falls. Look for ‘I think’ and ‘I believe’ statements.	When providing definitions about falls	'I think they [falls] are an unsolved problem.’

• Structural coding can be useful if a structured interview guide or focus group guide has been used by the researcher and the researcher stays close to the wording of the question and its prompts. The structured question is the structural code in the codebook, and the text segment should include the participant’s response and any dialogue following the question. Of course, this form of coding can be used even if the researcher does not follow a structured guide, which is often the reality of qualitative data collection. The relevant text segments are coded for the specific structure, as appropriate.³
• Content coding is informed by the research question(s) and the questions informing the analysis. The segmented text is grouped in different ways to explore relationships, hierarchies, descriptions and explanations of events, similarities, differences and consequences. The content of the text segment should be read and re-read to identify patterns and meaning, with the generated codes added to the codebook.
• Themes vary in scope, yet at the core they are phrases or statements that explain the meaning of the text. Researchers need to be aware that themes are considered a higher conceptual level than codes, and therefore should not be comprised of single words or labels. Typically, multiple codes will lead to a theme. Revisiting the research and analysis questions will assist the researcher to identify themes. Through the coding process, the researcher actively searches the data for themes. Examples of how themes may be identified include the repetition of concepts within and across transcripts, the use of metaphors and analogies, key phrases and common phrases used in an unfamiliar way.³

Framework analysis

This method originated in the 1980s in social policy research. Framework analysis is suited to research seeking to answer specific questions about a problem or issue, within a limited time frame and with homogenous data (in topics, concepts and participants); multiple researchers are usually involved in the coding process.^4-6 The process of framework analysis is methodical and suits large data sets, hence is attractive to quantitative researchers and health services researchers. Framework analysis is useful for multidisciplinary teams in which not all members are familiar with qualitative analysis. Framework analysis does not seek to generate theory and is not aligned with any particular epistemological, philosophical or theoretical approach.⁵ The output of framework analysis is a matrix with rows (cases), columns (codes) and cells of summarised data that enables researchers to analyse the data case by case and code by code. The case is usually an individual interview, or it can be a defined group or organisation.⁵

The process for conducting framework analysis is as follows⁵:

1. Transcription – usually verbatim transcription of the interview.

2. Familiarisation with the interview – reading the transcript and listening to the audio recording (particularly if the researcher doing the analysis did not conduct the interview) can assist in the interpretation of the data. Notes on analytical observations, thoughts and impressions are made in the margins of the transcript during this stage.

3. Coding – completed in a line-by-line method by at least two researchers from different disciplines (or with a patient or public involvement representative), where possible. Coding can be both deductive – (using a theory or specific topics relevant to the project – or inductive, whereby open coding is applied to elements such as behaviours, incidents, values, attitudes, beliefs, emotions and participant reactions. All data is coded.

4. Developing a working analytical framework – codes are collated and organised into categories, to create a structure for summarising or reducing the data.

5. Applying the analytical framework – indexing the remaining transcripts by using the categories and codes of the analytical framework.

6. Charting data into the framework matrix – summarising the data by category and from each transcript into the framework matrix, which is a spreadsheet with numbered cells in which summarised data are entered by codes (columns) and cases (rows). Charting needs to balance the reduction of data to a manageable few lines and retention of the meaning and ‘feel’ of the participant. References to illustrative quotes should be included.

7. Interpreting the data – using the framework matrix and notes taken throughout the analysis process to interpret meaning, in collaboration with team members, including lay and clinical members.

Reflexive thematic analysis

This is the thematic analysis approach developed by Braun and Clarke in 2006 and explained in the highly cited article ‘Using thematic analysis in psychology’.⁷ Reflexive thematic analysis recognises the subjectiveness of the analysis process, and that codes and themes are actively generated by the researcher. Hence, themes and codes are influenced by the researcher’s values, skills and experiences.⁸ Reflexive thematic analysis ‘exists at the intersection of the researcher, the dataset and the various contexts of interpretation’.^{9(line 5-6)} In this method, the coding process is less structured and more organic than in applied thematic analysis. Braun and Clarke have been critical of the use of the term ‘emerging themes’, which many researchers use to indicate that the theme was data-driven, as opposed to a deductive approach:

This language suggests that meaning is self evident and somehow ‘within’ the data waiting to be revealed, and that the researcher is a neutral conduit for the revelation of said meaning. In contrast, we conceptualise analysis as a situated and interactive process, reflecting both the data, the positionality of the researcher, and the context of the research itself… it is disingenuous to evoke a process whereby themes simply emerge, instead of being active co-productions on the part of the researcher, the data/participants and context.^10(p15)

Since 2006, Braun and Clarke have published extensively on reflexive thematic analysis, including a methodological paper comparing reflexive thematic analysis with other approaches to qualitative analysis,⁸ and have provided resources on their website to support researchers and students.⁹ There are many ways to conduct reflexive thematic analysis, but the six main steps in the method are outlined following.⁹ Note that this is not a linear, prescriptive or rule-based process, but rather an approach to guide researchers in systematically and robustly exploring their data.

1.  Familiarisation with data – involves reading and re-reading transcripts so that the researcher is immersed in the data. The researcher makes notes on their initial observations, interpretations and insights for both the individual transcripts and across all the transcripts or data sources.

2.  Coding – the process of applying succinct labels (codes) to the data in a way that captures the meaning and characteristics of the data relevant to the research question. The entire data set is coded in numerous rounds; however, unlike line-by-line coding in grounded theory (Chapter 27), or data segmentation in applied thematic analysis, not all sections of data need to be coded.⁸ After a few rounds of coding, the codes are collated and relevant data is extracted.

3.  Generating initial themes – using the collated codes and extracted data, the researcher identifies patterns of meaning (initial or potential themes). The researcher then revisits codes and the data to extract relevant data for the initial themes, to examine the viability of the theme.

4.  Developing and reviewing themes – checking the initial themes against codes and the entire data set to assess whether it captures the ‘story’ of the data and addresses the research question. During this step, the themes are often reworked by combining, splitting or discarding. For reflexive thematic analysis, a theme is defined as a ‘pattern of shared meaning underpinned by a central concept or idea’.^8(p39)

5.  Refining, defining and naming themes – developing the scope and boundaries of the theme, creating the story of the theme and applying an informative name for the theme.

6.  Writing up – is a key part of the analysis and involves writing the narrative of the themes, embedding the data and providing the contextual basis for the themes in the literature.

Themes versus codes

As described above, themes are informed by codes, and themes are defined at a conceptually higher level than codes. Themes are broader categorisations that tend to describe or explain the topic or concept. Themes need to extend beyond the code and are typically statements that can stand alone to describe and/or explain the data. Fereday and Muir-Cochrane explain this development from code to theme in Table 22.2.¹¹

Table 22.2. Corroborating and legitimating coded themes to identify second-order themes

First-order theme	Clustered themes	Second-order themes
The relationship between the source and recipient is important for feedback credibility, including frequency of contact, respect and trust The source of the feedback must demonstrate an understanding of the situational context surrounding the feedback message. Feedback should be gathered from a variety of sources. Verbal feedback is preferred to formal assessment, due to timing, and the opportunity to discuss issues.	Theme 1: Familiarity with a person increases the credibility of the feedback message. Theme 2: Feedback requires a situational-context. Theme 3: Verbal feedback is preferred over written feedback. Theme 4: Trust and respect between the source and recipient of feedback enhances the feedback message. Theme 5: Familiarity within relationships is potentially detrimental to the feedback process.	Familiarity When relationships enhance the relevance of feedback

_{*Note: This table is from an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.}

 

 

Table 22.3. Examples of thematic analysis

Title	Connection, constraint, and coping: a qualitative study of experiences of loneliness during the COVID-19 lockdown in the UK14	‘It is this very knowledge that makes us doctors’: an applied thematic analysis of how medical students perceive the relevance of biomedical science knowledge to clinical medicine15	Misconceptions and the acceptance of evidence-based nonsurgical interventions for knee osteoarthritis. a qualitative study16
CC Licence	CC BY 4.0	CC BY 4.0	Public Domain Mark 1.0
First author and year	McKenna-Plumley, 2021	Dickinson, 2020	Bunzli, 2019
Aim/research question	What are people’s experiences of loneliness while practising physical distancing due to a global pandemic? (p3)	‘To explore how medical students in their first clerkship year perceive the relevance of biomedical science knowledge to clinical medicine with the goal of providing insights relevant to curricular reform efforts that impact how the biomedical sciences are taught’ (p2)	‘To investigate the patient-related cognitive factors (beliefs/attitudes toward knee osteoarthritis and its treatment) and health system-related factors (access, referral pathways) known to influence treatment decisions.’ ‘Exploring why patients may feel that nonsurgical interventions are of little value in the treatment of knee osteoarthritis.’ (p1976)
Data collection	Semi-structured interviews by phone or videoconferencing software. Interview topics covered social isolation, social connection, loneliness and coping. Interview guide (supplementary file 2)14	55 student essays in response to the prompt: ‘How is biomedical science knowledge relevant to clinical medicine?’ A reflective writing assignment based on the principles of Kolb experiential learning model	Face-to-face or phone interviews with 27 patients who were on a waiting list for total knee arthroplasty.
Thematic analysis approach	Reflexive thematic analysis	Applied thematic analysis	Framework analysis
Results	Table of themes and illustrative quotes: 1. Loss of in-person interaction causing loneliness 2. Constrained freedom 3. Challenging emotions 4. Coping with loneliness	1. Knowledge-to-practice medicine 2. Lifelong learning 3. Physician-patient relationship       4. Learning perception of self	Identity beliefs – knee osteoarthritis is ‘bone on bone’ Casual belief – ‘osteoarthritis is due to excessive loading through the knee’ Consequence beliefs – fear of falling and damaging the joint Timeline beliefs – osteoarthritis as a downward trajectory, the urgency to do something and arriving at the end of the road.

Advantages and challenges of thematic analysis

Thematic analysis is flexible and can be used to analyse small and large data sets with homogenous and heterogenous samples. Thematic analysis can be applied to any type of data source, from interviews and focus groups to diary entries and online discussion forums.¹ Applied thematic analysis and framework analysis are accessible approaches for non-qualitative researchers or beginner researchers. However, the flexibility and accessibility of thematic analysis can lead to limitations and challenges when thematic analysis is misapplied or done poorly. Thematic analysis can be more descriptive than interpretive if not properly anchored in a theoretical framework.¹ For framework analysis, the spreadsheet matrix output can lead to quantitative researchers inappropriately quantifying the qualitative data. Therefore, training and support from a qualitative researcher with the appropriate expertise can help to ensure that the interpretation of the data is meaningful.⁵

Summary

Thematic analysis is a family of analysis techniques that are flexible and inductive and involve the generation of codes and themes. There are three main types of thematic analysis: applied thematic analysis, framework analysis and reflexive thematic analysis. These approaches span from structured coding to organic and unstructured coding for theme development. The choice of approach should be guided by the research question, the research design and the available resources and skills of the researcher and team.

 

References

1. Clarke V, Braun V. Thematic analysis. J Posit Psychol. 2017;12(3):297-298. doi:10.1080/17439760.2016.1262613
2. Guest G, MacQueen KM, Namey EE. Introduction to applied thematic analysis. In: Guest G, MacQueen, K.M., Namey, E.E., ed. Applied thematic analysis. SAGE Publications, Inc.; 2014. Accessed September 18, 2023. https://methods.sagepub.com/book/applied-thematic-analysis
3. Guest G, MacQueen, K.M., Namey, E.E.,. Themes and Codes. In: Guest G, MacQueen, K.M., Namey, E.E., ed. Applied thematic analysis. SAGE Publications, Inc.; 2014. Accessed September 18, 2023. https://methods.sagepub.com/book/applied-thematic-analysis
4. Srivastava A, Thomson SB. Framework analysis: A qualitative methodology for applied policy research. Journal of Administration and Governance. 2009;72(3). Accessed September 14, 2023. https://ssrn.com/abstract=2760705
5. Gale NK, Heath G, Cameron E, Rashid S, Redwood S. Using the framework method for the analysis of qualitative data in multi-disciplinary health research. BMC Med Res Methodol. 2013;13:117. doi:10.1186/1471-2288-13-117
6. Smith J, Firth J. Qualitative data analysis: the framework approach. Nurse Res. 2011;18(2):52-62. doi:10.7748/nr2011.01.18.2.52.c8284
7. Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol. 2006;3(2):77-101. doi:10.1191/1478088706qp063oa
8. Braun V, Clarke V. Can I use TA? Should I use TA? Should I not use TA? Comparing reflexive thematic analysis and other pattern-based qualitative analytic approaches. Couns Psychother Res. 2021;21(1):37-47. doi:10.1002/capr.12360
9. Braun V, Clarke V. Thematic analysis. University of Auckland. Accessed September 18, 2023. https://www.thematicanalysis.net/
10. Braun V, Clarke V. Answers to frequently asked questions about thematic analysis. University of Auckland. Accessed September 18, 2023. https://www.thematicanalysis.net/faqs/
11. Fereday J, Muir-Cochrane E. Demonstrating Rigour Using Thematic Analysis: A Hybrid Approach of Inductive and Deductive Coding and Theme Development. International Journal of Qualitative Methods. 2006;5(1):80-92. doi: 10.1177/160940690600500107
12. Ayton D, Manderson L, Smith BJ. Barriers and challenges affecting the contemporary church’s engagement in health promotion. Health Promot J Austr. 2017;28(1):52-58. doi:10.1071/HE15037
13. Ayton D, Watson E, Betts JM, et al. Implementation of an antimicrobial stewardship program in the Australian private hospital system: qualitative study of attitudes to antimicrobial resistance and antimicrobial stewardship. BMC Health Serv Res. 2022;22(1):1554. doi:10.1186/s12913-022-08938-8
14. McKenna-Plumley PE, Graham-Wisener L, Berry E, Groarke JM. Connection, constraint, and coping: A qualitative study of experiences of loneliness during the COVID-19 lockdown in the UK. PLoS One. 2021;16(10):e0258344. doi:10.1371/journal.pone.0258344
15. Dickinson BL, Gibson K, VanDerKolk K, et al. “It is this very knowledge that makes us doctors”: an applied thematic analysis of how medical students perceive the relevance of biomedical science knowledge to clinical medicine. BMC Med Educ. 2020;20(1):356. doi:10.1186/s12909-020-02251-w
16. Bunzli S, O’Brien P, Ayton D, et al. Misconceptions and the acceptance of evidence-based nonsurgical interventions for knee osteoarthritis. A Qualitative Study. Clin Orthop Relat Res. 2019;477(9):1975-1983. doi:10.1097/CORR.0000000000000784