Tips on transcribing qualitative interviews and naturally-occurring interaction

After interviews have been conducted or events have been recorded, the task of transcription begins. For those who have not transcribed before, it is easy to under-estimate the amount of time needed to transcribe interviews and interaction. What does transcription entail? In my experience, it takes at least an hour to transcribe every 15 minutes of interview talk. If I am transcribing interaction that involves multiple speakers, or if I am using more detailed conventions to record other features of talk, transcribing takes much longer. Of course, the time spent on transcribing will also vary depending on the transcriber’s typing speed, the speed of talk, the number of speakers involved, the technologies used to transcribe, and what is made available to transcribe by the recording. For example, video recordings provide much more information that may be transcribed, and can take much longer to transcribe than audio-recordings if descriptions of actions and activities are included in transcripts.

There are many tools to assist with transcription.

I use Express Scribe

but you might also like to look at

Inqscribe

Transana

Audacity

Whatever approach I am using to transcribe, I routinely begin each transcription by including the following information:

• Name of the research project
• Speakers (identified by pseudonyms)
• Date and time of the interview or recording of the event
• Context and setting
• Audio file name or number
• Duration of interview or recorded event

If there is a particular point in a recording that I plan to return to repeatedly, I include time-stamps so I can locate that point in the audio-file quickly. If you are transcribing interaction that has been associated with at Qualitative Data Analysis software (QDAS), such as NVivo, it is also possible to synchronize the recording with the transcript. This will be invaluable if you need to return to particular points in the data to listen to the interaction as you are working on analyzing data. If there is a specific sequence that I intend to use in presentations — in cases in which I have gained informed consent to do so — I might copy and edit the audio-file so I can easily access the shorter sequence that will be used.

In cases in which transcripts will be printed for review, I include automatic line numbering. When transcriptions will be imported into a QDAS program, I don’t insert line numbering, since it is not necessary. I use line numbers on printed transcripts because they provide an easy way to identify lines should I be annotating transcriptions or working with other project members on data analysis.

How does one transcribe? There is a good deal more decision-making involved in transcription then one might initially think. For example, questions that arise in the process of transcription include:

• How does one transform oral talk into sentences?
• What should a transcribed do about dialect? Contractions (such as “dunno” or “gonna”?
• Should laughter or other kinds of interactions (e.g., crying) be included in the transcription? How should that be notated?
• What about re-starts and verbal “tics” (e.g., “like” “you know” “mm” and “uh” to name but a few)?
• What happens when participants mention places and names related to a study in the transcription? Should these be included?

Linguistic anthropologist, Elinor Ochs (1979) in a chapter entitled “Transcription as theory,” argues that the process of transcription is theoretically-informed. Any particular transcription is imbued with assumptions about what the data means, how it might be analyzed, and what kinds of claims might be made. Thus, there are many different formats for transcription. For example conversation analysts transcribe silences, restarts, overlapping talk, and other paralinguistic features of talk (for a guide to applying CA’s transcription conventions, see Liddicoat, 2007).

More commonly seen in qualitative research is an approach that values the topics of talk without acknowledgement of how the talk is produced. At face value, it might seem that the “topic” is the most important feature of talk. Yet, one only needs to think about a comment delivered in an ironic or sarcastic manner to recognize that “how” topics of talk are delivered is an important feature of talk that is used by listeners to interpret what has been said by speakers. Fortunately, there are plenty of resources to assist with decision making concerning what and how to transcribe interaction. For anyone beginning to transcribe Poland (2002) provides useful pointers to consider. For those wanting to complete more detailed transcriptions Jenks (2011) is helpful. If you are working with video interaction Heath, Hindmarsh, and Luff (2010) provide useful information on how to collect, transcribe, analyze video data. Whatever method one uses, however, thought needs to be given to how one plans to analyze data, since how data are transcribed relates to what approaches to analysis might be used and what kinds of research questions might be asked of data.

Further challenges await scholars who are presenting findings from studies in which interviews were conducted in another language. Bilingual and multilingual speakers are well aware of the difficulties of translating words and concepts from one language to another. Qualitative researchers using translated data need to account for how interviews have been translated, at what point, and how analysis was accomplished. For more information on using translated data in qualitative inquiry, see Regmi, Naidoo, and Pilkington (2010); Temple and Edwards (2002); Temple and Young (2004); and Ten Have (2007). Ten Have provides examples of how transcriptions can include both the original language with the translation.

This short introduction to transcribing qualitative data – whether naturally occurring interaction or interview data, brings to the fore some of the issues involved, as well as some of the resources available to assist. Some scholars have suggested that with increasing use of audio- and video-recordings, transcription will no longer be necessary. In my view, transcriptions of interaction are still a necessary step needed in the analytic process. A written record of interaction provides me with a document that I return to repeatedly to re-read and reflect on. Listening to audio-recordings and video-recordings is also a way to become acquainted with data.

For newcomers to transcription, it can sometimes seem that this is a tedious task best left to others. Another way to view transcription is that it is part of the analytic process. Listening carefully to what people say in interviews and naturally occurring settings is the first step to meaning-making and interpretation. By valuing this step of the research process, and undertaking careful and thorough transcription, qualitative researchers will come to know their data intimately. This enhances analytic sensibilities and helps us to think about and formulate findings and research reports.

I’ve mentioned some resources on transcription here… there is much more writing on this topic. For those interested in learning more, see transcription-references_qualpage

Best wishes with that work.

Kathy R.

References

Heath, C., Hindmarsh, J., & Luff, P. (2010). Video in qualitative research: Analysing social interaction in everyday life. Los Angeles, London, New Delhi, Singapore, Washington DC: Sage.

Jenks, C. J. (2011). Transcribing talk and interaction. Amsterdam/Philadelphia: John Benjamins Publishing Company

Liddicoat, A. J. (2007). An introduction to conversation analysis. London: Continuum.

Ochs, E. (1979). Transcription as theory. In E. Ochs & B. Shieffelin (Eds.), Developmental pragmatics (pp. 43-72). New York: Academic Press.

Poland, B. D. (2002). Transcription quality. In J. Gubrium & J. A. Holstein (Eds.), Handbook of interview research: Context and method (pp. 629-650). Thousand Oaks: Sage.

Regmi, K., Naidoo, J., & Pilkington, P. (2010). Understanding the processes of translation and transliteration in qualitative research. International Journal of Qualitative Methods, 9(1), 16-26.

Temple, B., & Edwards, R. (2002). Interpreters/translators and cross-language research:Reflexivity and border crossings. International Journal of Qualitative Methods, 1(2), 1-22.

Temple, B., & Young, A. (2004). Qualitative research and translation dilemmas. Qualitative Research, 4(2), 161-178. doi:10.1177/1468794104044430

Ten Have, P. (2007). Doing conversation analysis: A practical guide (2nd ed.). London, UK: Sage.