Thematic Analysis Presentation PPTS.pptx

Thematic Analysis

1. Thematic Analysis © Virginia Braun, Victoria Clarke

2. Transcribing audio data for thematic analysis Virginia Braun and Victoria Clarke

3. Overview Transcription is not a straightforward process of translation, but an active process of transformation. Factors that contribute to (orthographic) transcript quality. Ethics and transcription. Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013) - also available as a handout. The transcription process. Transcription exercise. 3

4. Transcription is not a straightforward process of translation(1 of 2) The point of transcription is to produce a written record of an audio recording – and that transcript usually serves as the foundation for analysis, the material you predominantly work with in coding, theme development and interpretation. But transcription is not straightforward. Whole books have been written about transcription – it is far from a simple ‘technical’ concern. 4

5. Transcription is not a straightforward process of translation(2 of 2) For a start, spoken and written language are very different. We speak in really messy ways, without ‘grammatical’ punctuation, and it’s often a jumble of pauses, repetition, and hesitations. We regularly change things like the pace, tone or volume of our speech. This means a transcript of speech looks nothing like written language – or what you might expect from dialogue as it appears in say a play, or a novel. For much qualitative research, and thematic analysis, it’s important the transcript captures ‘real’ speech – what was said and (to a greater or lesser extent) how it was said. 5

6. Transcription is an active process of transformation(1 of 2) Transcription is better understood as a process of transformation rather than replication. Sandelowski (1994 ) suggested that transcription turns “raw” data (for example, audio recordings) into “partially cooked” data (the transcript thereof) – some processing of the data and change has occurred. The process of transcription involves a range of decisions about what information to include and what to exclude, and things are lost and gained in the process. 6

7. Transcription is an active process of transformation(2 of 2) Transcription, then, is an active process, and it’s important to treat it as such, rather than a neutral technology to deliver information into a more easily analysable form. The most common form of transcription for thematic analysis is ‘orthographic’ or verbatim, which focuses primarily on producing a ‘complete’ record of the spoken words, including partial words or stutters, along with some other features such as laughter. 7

8. Factors that contribute to (orthographic) transcript quality(1 of 6) Using a clear, well-defined and systematic notation system. With orthographic transcription, people often develop their own idiosyncratic approach, but there is value in following guidelines (such as those we described in Successful Qualitative Research ; Braun and Clarke, 2013). Including page or line numbers. 8

9. Factors that contribute to (orthographic) transcript quality(2 of 5) Including brief contextual information at the start such as the participant number or code, the date the interview or focus group was conducted, the name of the interviewer, important demographic characteristics; this can be especially useful if sharing transcripts with a supervisor or when working in a research team (but be mindful of the importance of participant anonymity). Providing a clear indication of who is speaking and what they are saying. 9

10. Factors that contribute to (orthographic) transcript quality(3 of 5) Providing a as complete and clear as possible record of all utterances, as they were uttered – do not ‘correct’ things like grammar, slang or spoken abbreviations. It is very easy to do some of this unintentionally – good transcription requires focused attention on what was actually spoken, instead of what we hear! 10

11. Factors that contribute to (orthographic) transcript quality(4 of 5) Consistent inclusion or exclusions of some ‘paralinguistic’ features of the data (for example, noting if a participant is crying, or laughing, or if there is hesitation). Not importing written language grammar into the transcript. Punctuation adds ‘meaning’ to transcripts (the placement of commas can change meaning – acknowledged in the title of the excellent book on punctuation Eats, shoots and leaves by Lynne Truss). This can be helpful OR unhelpful – therefore, do not include it initially, but listen back to the audio to check, and add in at a later point. 11

12. Factors that contribute to (orthographic) transcript quality(5 of 5) ‘Accuracy’ – it’s really common to not be able to hear something clearly, or to mishear something, and this can affect the analysis. Transcription errors are neither uncommon nor inconsequential (Poland, 2002). A good transcript will indicate where something isn’t decipherable, or it is a ‘best guess’ as to what was spoken. Readability – aided through the use of new lines for different speakers* and the use of hanging indent system, to clearly identify each time a new speaker starts. * sometimes single interjections such as “yes” or “ mhm ” – especially when by an interviewer – may better be transcribed in parentheses with the flow of a longer speech attributed to one person.   12

13. Ethics and transcription(1 of 2) There are ethical concern related to transcripts – and transcription, because data are usually confidential. As transcripts usually contain potentially identifiable data, it is usually recommended to password protect or encrypt the files, and use an anonymised version for analysis. If an external transcriber is used, they’ll need to sign a confidentiality agreement. (You should also consider the nature of the data content; some topics are distressing for people and it’s ethically important they make informed choices about vulnerability.) 13

14. Ethics and transcription(2 of 2) If an transcription app or programme is used, check what they do in relation to cloud storage of files, etcetera, to ensure there is no breach of ethical and data protection requirements (and include this detail in an ethics application). 14

15. Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(1 of 10) Feature Notation The identity of the speaker Use the speaker’s name followed by a colon (for example, Anna:[colon ] ) to signal the identity of a speaker or Moderator:[colon]/Mod:[colon] or Interviewer:[colon]/Int:[colon] for when the moderator/interviewer is speaking; it may also be appropriate to use the moderator/interviewer’s first name. Turn-taking Start a new line every time a new speaker enters the conversation. 15

16. Feature Notation Laughing, coughing etcetera ((laughs)) and ((coughs)) signals a speaker laughing or coughing during a turn of talk; ((General laughter)) signals multiple speakers laughing at once and should be appear on a separate line (to signal that no one speaker ‘owns’ the laughter). Pausing ((pause)) signals a significant pause (in other words, a few seconds or more; precise timings of pauses are not necessary); can also use (.) to signal a short pause (a second or less) or ((long pause)) to signal a much longer pause. 16 Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(2 of 10)

17. Feature Notation Spoken abbreviations (for example, TV for television; WHO for World Health Organisation) If someone speaks an abbreviation then use that abbreviation (for example, TV), but do not abbreviate unless a speaker does so. Overlapping speech Type ((in overlap)) before the start of the overlapping speech. Inaudible speech ((inaudible)) for speech and sounds that are completely inaudible; when you can hear something but you’re not sure if it’s correct use brackets (ways of life) to signal transcriber doubt or (ways of life/married wife) to signal multiple potential hearings. 17 Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(3 of 10)

18. Feature Notation Uncertainty about who is speaking ?[question mark] to signal uncertainty about the speaker or name followed by a question mark (for example, Judy?[question mark]) to signal your best guess. Or, if you can only identify the sex/gender of the speaker, F?[question mark] or M?[question mark] to signal that the speaker is a woman or a man, but be mindful that gender isn’t always easily readable from a person’s voice. 18 Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(4 of 10)

19. Feature Notation Non-verbal utterances Render phonetically and consistently. Common non-verbal sounds uttered by English as a first language speakers include ‘erm’, ‘er’, ‘mm’, ‘ mmhm ’. Spoken numbers Spell out all numbers and be mindful of the difference between ‘a hundred’ and ‘one hundred’. 19 Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(5 of 10)

20. Feature Notation Do I use punctuation? It is common to use punctuation to signal some features of spoken language such as using a question mark to signal the rising intonation of a question and a comma to signal the slight pause or emphasis of continuing intonation in speech. However, adding punctuation to a transcript is not straightforward and it is important to be mindful of the ways in which adding punctuation can change the meaning of an extract of data. Equally, punctuation enhances the readability of spoken data, especially extracts quoted in written reports. 20 Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(6 of 10)

21. Feature Notation Cut off speech and sounds This level of detail is not necessary for most experiential forms of analysis, although it can be useful to signal moments when participants are struggling to articulate their thoughts, feelings etcetera. and cut off speech can be a part of this – to signal cut off speech type out the sounds you can hear then add a dash (for example, wor -[dash]). Emphasis on particular words Again, this level of detail is not necessary for most experiential forms of analysis, although it can be useful to signal words or sounds that are particularly emphasised by underlining (for example, word[underlined] ). 21 Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(7 of 10)

22. Feature Notation Reported speech When a speaker reports the speech (or thoughts) of another person (or reports their own speech in the past) signal this with the use of inverted commas around the reported speech. For example, and she said [inverted comma open]‘I think your bum does look big in that dress’[inverted comma close] and I said [inverted comma open]‘thanks a bunch’[inverted comma close]…. 22 Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(8 of 10)

23. Feature Notation Accents and abbreviations/ vernacular usage It’s important not to transform participants’ speech into ‘standard’ English; however, fully representing a strong regional accent can be a complex and time consuming process! A good compromise is to signal only the very obvious or common (and easy to translate into written text) abbreviations and vernacular usage such as ‘cos’ instead of ‘because’ or a Welsh speaker saying ‘me Mam’ (instead of the English ‘my Mum’), unless it is absolutely critical for your analysis to fully represent exactly how a speaker pronounces words and sounds. 23 Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(9 of 10)

24. Feature Notation Names of media These should be presented in italics (for example, The Walking Dead , Men’s Health [italicised]). Identifying information You can either change identifying information such as people’s names and occupations, places, events etcetera (for example, ‘Bristol’ to ‘Manchester’) or insert generic descriptions of the information in square brackets (for example, [square bracket open][brother’s name][square bracket closed). 24 Transcription notation system for orthographic transcription (revised from Braun and Clarke, 2013 )(10 of 10)

25. The transcription process (‘old school ’)(1 of 2) The process of transcription itself is fairly straightforward, once you have the equipment you need. The only essential item is a digital playback programme (assuming a digital recording), preferably one that is designed for transcription (for example, the free basic software Express Scribe ), because they often have features like speed adjustment, which can really help with utterances that are hard to decipher, and ‘auto-reverse’ features, whenever you pause. 25

26. The transcription process (‘old school ’)(2 of 2) Alongside that, a transcription foot pedal can be useful – programmes like Express Scribe work with such pedals – because they leave your hands free for typing! The other item that can be invaluable is good quality headphones – as well as removing extraneous noise that impacts hearing and concentration, headphones can aid in hearing sounds, facilitating focus and clarity. 26

27. The transcription process(1 of 2) Once you’re set up and ready, with a Word file open, start playing the data. Type everything you hear. Focus on the words/sounds, not the meaning, because it can facilitate ‘accuracy’. You can only usually transcribe a short segment at a time, and transcription quality requires a number of repeat listens to each segment you work on, to ensure it is clear and ‘accurate’. This means there is lots of stopping and starting, and moving back and forth through the recording and transcript. 27

28. The transcription process(2 of 2) As well as working to get a full and complete transcript through this process, it also always pays to listen again to the full recording, after completion – and a little break from it! You’ll likely catch absences and errors in what’s there, and even catch words you couldn’t make out before. If someone else has done the transcribing for you – this ‘review’ phase is vital. Transcripts usually contain potentially identifiable data, so we recommend password protecting or encrypting them, and using an anonymised version (see Braun and Clarke, 2013) for analysis. 28

29. Transcription software(1 of 2) Various options for voice recognition transcription are now available, from those that ‘transcribe’ direct from a digital file upload to those that operate more like ‘dictation’ – and transcribe from playback. One option is Google Docs voice typing, where you read aloud the interview as it plays in your headphones. People report different experiences of this. For some, this can cut transcription time in half compared to typing. But for others, it can be more time-consuming to do this and then check the transcript; and also time is lost because you are disengaged from the data, compared to transcribing it yourself. 29

30. Transcription software(2 of 2) Otter.ai transcription software is often recommended. The ‘free’ subscription version allows six-hundred minutes of transcription per month; the (pay) premium version gives added features, such as exporting direct to Word . To give a sense of functionality (as of two-thousand-and-twenty), even with very different accents, it can mix up speakers quite a bit, and identify three different speakers when there are only two; it can make some comedy word errors, but overall provides a very good base for editing and developing a quality transcript. 30

31. Transcription software - the importance of the check and edit The most important aspect to emphasise with automated transcription (the same goes with transcripts done by someone else, and even yourself, to a different extent), is: check and edit ! A digitally-produced transcript will need to be fully revised and edited, not just for formatting but for content . 31

32. Transcription exercise(1 of 2) Have a go at transcription, if you are planning to use audio recorded data. Download Express Scribe (basic). Find an audio or video file of fairly ‘natural’ talking (for example, not something like a Ted talk), preferably a conversation. Some of the conversational-style podcasts, for instance, would work well. Have a go at transcribing about five minutes of the dialogue, using the guidelines provided. 32

33. Transcription exercise(2 of 2) Once you have transcribed the talk, reflect on: a) How hard or easy the activity was, and in what ways; and b) How long it took. What factors impacted how easy or hard you found this experience? After a break, go back and listen to the audio and read the transcript again. Do you pick up more information? Did you miss anything originally? Did you make any errors ? 33

34. References Braun, V., & Clarke, V. (2013). Successful qualitative research: A practice guide for beginners . London: Sage. Poland, B. D. (2002). Transcription quality. In J. F. Gubrium & J. A. Holstein (Eds.), Handbook of Interview Research: Context & Method (pp. 629-649). Thousand Oaks, CA: Sage. Sandelowski, M. (1994). Focus on qualitative methods: Notes on transcription. Research in Nursing & Health , 17, 311-314. 34