To transcribe or not to transcribe – that is the question!

Our first blog post by Dr Sue Nielsen :) This is a regular topic of conversation for us and the researchers we work with. Let us know what you think...

Qualitative researchers are often expected to make exhaustive written transcriptions of audio and video recordings of interviews, observations and other qualitative data. However, exhaustive transcribing is very costly, often poorly done and may delay the progress of the research project.

Researchers need to consider the following issues. Cost and skill – Most researchers are not fast, accurate transcribers. On the other hand, qualified transcribers may not understand the domain, leading to problems with professional and discipline terminology, jargon etc.

What and how - Transcribing is a research activity not a technical process. Most qualitative researchers analyse as they listen to or look at their data. What and how to transcribe depends on the purpose of the researcher. Most researchers need to ‘reduce’ and synthesise their data so full transcriptions are unnecessary. Different research methods such as conversation analysis, grounded theory and thematic analysis require very different skills and approaches.

Accuracy - Transcribing audio and visual data may lose authenticity and relevant information. Capturing emotional, vocal, spatial and other data is very difficult and time consuming. Much of this data gets lost or forgotten and will not be taken into account in the data analysis phase.

Trustworthiness - The purpose of the qualitative research project is to write up the findings. To ensure trustworthiness, many authors suggest establishing an audit trail to enable readers or examiners to follow the line of inquiry and analysis from the interpretation back to the original data. Researchers should keep descriptions and registers of data collected, as well as how the data was handled in the analysis process. It is misleading to assume that transcripts alone will accurately represent the original data.

Help - Computer based programmes can assist in addressing these issues. Small projects may simply use word processing, spreadsheets or databases. Longer and more complex projects will benefit from the use of a CAQDAS programme. Programmes such as NVivo enable researchers to make notes on and code audio and visual data alongside textual data.

Whatever the size of the project, think carefully before embarking on costly and possibly misleading transcriptions of all your data.

The following papers provide a good argument for working with data in its original state and an overview of emerging technologies which support data analysis and coding. Markle, D.T., West, R.E., & Rich, P.J (2011) Beyond Transcription: Technology, Change, and Refinement of Method, FQS Forum: Qualitative Research, 12 (3) September, Article 21. http://www.qualitative-research.net/index.php/fqs/article/view/1564/3249

McLellan, E., McQueen, K., & Neidig, J. (2003). Beyond the qualitative interview: Data preparation and transcription. Field Methods, 15(1), 63-84.

Love to hear your comments!