Toolkit 08: Transcribing your own qualitative data

Hazel Burke, Lisa Jenkins and Victoria Higham, May 2010

Summary

The aim of this toolkit is to help you plan properly for transcribing, and save you time by helping you do your transcribing more efficiently.

This toolkit deals with transcribing your own, or your team’s, data. If you are considering sending your data to somebody else or an external transcribing agency, see our toolkit on outsourcing transcribing (coming soon).

The toolkit includes advice on project planning for transcribing, tips for good quality recordings, advice on equipment and software and suggestions to help you save time while transcribing. Plus you can also use our spreadsheet to work out how long your transcribing will take and watch our mini-tutorial to show you how to add line numbers to your transcript.

Mini-tutorial showing how to add line numbers to your transcript (about 1 minute):

In Word, click on the first line you want to be numbered, then go to File > Page Set up > Layout tab - there is a drop-down box called 'Apply to' where you can select 'this point forward'.

Some comments on transcribing recordings using voice recognition software

Voice recognition software, such as Dragon Naturally Speaking or MacSpeech Scribe, is designed to turn speech into text on your computer. Though this sounds like a panacea for transcribing, there are some hitches connected with it. Mostly, these are to do with the fact that the software needs to be 'trained' to recognise a particular voice before it will work. As most interview/focus group data consists of lots of different voices this causes problems for the software.

However, a colleague has used voice recognition software to help with transcribing, as follows.

  1. Install a microphone and the voice recognition software on your computer and train the software to recognise your voice.
  2. Load your interview recording onto a laptop and plug in a pair of headphones. (You could use your digital recorder but this might give less accurate play/pause/rewind options.)
  3. Listen to the interview recording, press pause, and then speak the section you have just listened to into your computer.

This is obviously still a lengthy process, but once you are used to it you might find it quicker than typing your transcripts (our colleague reckons it takes around 4-5 hours per hour of interview). If you have problems using a keyboard, or if you type slowly, then this method might be worth a try!

If you have any other tips for working with voice recognition software for transcribing then do let us know so we can update our toolkit.