Transcription

Achieving high-quality data transcriptions

Transcription is a translation between forms of data. In the social sciences, this is most commonly converting audio recordings of interviews or discussions to text format.

Whilst audio-transcription is often part of the analysis process, it also enhances the sharing and reuse potential of qualitative research data. It is recommended that researchers make transcriptions of interviews; full-transcriptions significantly extend the potential for analysis and the re-use of a research collection, both by the original researchers and by secondary users. 

If interviews are not transcribed, then recorded interviews could be archived alongside summaries, but it may be difficult or time-consuming to effectively anonymise the audio files.

High quality and consistent transcription that matches the analytic and methodological aims of the research is part of good data management planning. Attention needs to be given to transcribing conventions needed for the research, transcription instructions or guidelines for transcribers, and a template to ensure uniformity across a collection.

How do I transcribe interviews?

Audio-visually recorded interviews are often transcribed manually. However, where audio recordings are of good quality with minimal background noise, it may also be possible to use automatic speech recognition (ASR) software to do an intial transcription, which is then further edited to ensure correct turn-taking, assign speaker tags, and checked for accuracy.

A standard transcription structure is recommended if transcripts are to be archived, or if Computer Assisted Qualitative Data Analysis (CAQDAS) software is to be used to analyse the data. Our templates below detail how to transcribe data:

Standard transcription template

It is vital to develop a standard transcription template for transcribers to follow. Provide written instructions or guidelines indicating the required transcription style, layout and editing, to ensure uniformity across the transcriptions.

A good rule of thumb is to make sure transcriptions feature the following:

  • A unique interview identifier, such as a name or number.
  • A uniform layout throughout a research project or data collection.
  • Speaker tags to indicate turn-taking or question/answer sequence in conversations.
  • Line breaks between turn-takes.
  • Numbered pages.
  • Header with brief interview or event details, such as date, place, interviewer name and interviewee details
  • Annotations to show clearly where amendments have been made, for example, translation to lay words or real names changed. Anonymisation edits are best placed between square brackets.

See the UK Data Service transcription template.

Format for transcription

Some best practice considerations for designing the format for transcription are listed below:

  • Consider the compatibility of your transcription format with the import features of qualitative data analysis software before developing a template or guidelines. Document headers and textual formatting, such as italics or bold, may be lost when transcripts are imported. Text formatted in two columns indicating speakers and utterances may also be problematic
  • Draft transcriber instructions or guidelines, indicating your required transcription style, layout and editing, especially if multiple transcribers carry out work;
  • Either anonymise data during transcription, or mark sensitive information for later anonymisation
  • Provide a translation or at least a summary of each interview in your national language or in English (if the text is not in your national language) in addition to transcriptions in the original language.

Automatic speech recognition (ASR)

Recent advances in automatic speech recognition (ASR) have made computer-assisted transcription a cost-effective and time-efficient option.

Various software packages designed to automatically transcribe text from an audio source are available. These softwares can require significant training and calibration to be able to recognise a particular voice, accent and dialect.

They can be useful if interviews are similar and avoid any peculiar jargon. It is best to use them with caution and always ensure checking of the whole text.

Data security

Attention should be paid to data security when transmitting recordings and transcripts that contain personal and confidential information to transcribers, and back to the researcher.

It is essential to establish data security procedures for the transcriber to follow when handling the data; a non-disclosure agreement can be drawn up with transcribers and files can be encrypted before transfer.

Transcription methods

Transcription methods depend very much upon your theoretical and methodological approach and can vary between disciplines. A thematic sociological research project usually requires a denaturalised approach, i.e. most like written language (Bucholtz, 2000) because the focus is on the content of what was said and the themes that emerge from that.

A project using conversation analysis would use a naturalised approach, i.e. most like speech, whereby a transcriber seeks to capture all the sounds they hear and use a range of symbols to represent particular features of speech in addition to the spoken words; for example representing the length of pauses, laughter, overlapping speech, turn-taking or intonation.

A psycho-social method transcript may include detailed notes on emotional reactions, physical orientation, body language, use of space, as well as psycho-dynamics in the relationship between the interviewer and interviewee.

Some transcribers may try to make a transcript look correct in grammar and punctuation, considerably changing the sense of flow and dynamics of the spoken interaction. Transcription should capture the essence of the spoken word, but need not go as far as the naturalised approach.

Reference: Bucholtz, M. (2000) The Politics of Transcription. Journal of Pragmatics 32: 1439-1465.

Must I transcribe interviews if I want to archive them?

It is recommended that transcriptions of interviews are made. Full transcriptions significantly extend the potential for analysis and re-use of a research collection, both by the original researchers and by secondary users. Transcription should be seen as a step within the analytical process of research, rather than as a mechanical conversion of data.

If interviews are not transcribed, then recorded interviews could be archived alongside summaries, but it may be difficult or time consuming to effectively anonymise the audio files. See guidance on transcription.