Options for transcribing your data in NVivo
15 April 2014 - IN audio, in, NVivo, recognition, TranscribeMe, transcribing, video, Voice
Transcription is a challenge most researchers will face at some stage in their qualitative journey - and if you're dealing with hours of recorded media it can be a thorny one.
While transcribing your own data can have significant analytical benefits (get close and personal with the material), it may not be practical for everyone - particularly if you're time-strapped, can't touch type or just have a low tedium threshold.
So, how does NVivo help you with the transcription conundrum?
I thought I'd layout a few scenarios - so you can see what's possible.
1. Let TranscribeMe do the hard work for you
The latest update to NVivo for Windows (Service Pack 5) gives you the option to have your media files professionally transcribed using TranscribeMe - a competitively priced, integrated transcription service.
While working away in NVivo, you can open the source you want transcribed (or select multiple sources in List View) and click the Purchase Transcript option:
You'll be prompted to sign-up for a TranscribeMe account - then you can upload the media (securely) and sit back while others do the work for you.
I like the fact that you can choose what sort of transcript you'd prefer; word-by-word verbatim, clean verbatim (removing 'hmm' and 'you know') or one with improved grammar and readability. The turnaround time is impressive too - you can choose from 24 to 72 hours.
The transcript is automatically downloaded into your project and NVivo will let you know when it's available. Then, the only thing you need to do is code the material and come up with brilliant insights. Easy.
To find out more about TranscribeMe (and the 25% discount valid until June 30) - head to the QSR Website.
2. Do it yourself using NVivo's transcribe features
If you decide to transcribe the material yourself - you can play the media in NVivo and transcribe it as you listen.
You can adjust the audio speed and use shortcut keys to make things easier - all of this is explained in Create audio or video transcripts in NVivo.
To really free things up, you might want to use a foot pedal - this one from Pedable is quite popular:
- Foot pedal for use with Pedable software
If the structure of conversation is important to you (maybe you're a sociolinguist or phenomenologist), then you'll probably want your interviews or focus groups fully transcribed - but if you're on a fact-finding mission (not really interested in the subtleties of expression) then maybe taking notes is all you need to do.
3. Say 'no' to transcripts
Since you can code your audio or video sources directly, maybe you don't need transcripts at all.
Consider this example:
- Listen to your interviews.
- Each time you hear participants talk about food additives, select the section on the timeline and code it at the node Food Additives.
- Open the node Food Additives and listen to all the coded snippets in one place.
- Make a memo to summarize your findings.
Depending on your research goals and methodology this approach may work for you. The only draw-back is that you can't use Text search or Word Frequency queries to find content in the interviews (although you can search and code the memos you make).
4. Take an experimental approach
Come on - this is 2014 - cars can drive themselves, Siri can explain the universe. So, why-oh-why can't my computer convert audio to text?
We are all waiting patiently for accurate voice-to-text technology. Products like Dragon NaturallySpeaking are getting there - as long as one person is speaking clearly (rarely the case in interviews and focus groups).
Having said that, one of our NVivo users is having success with the following approach:
- Listen to interview with headphones
- Dictate what you hear into Dragon NaturallySpeaking
- Import the transcript into NVivo.
Sounds tricky but might be worth exploring - you could follow this discussion on LinkedIn (NVivo Users Group).
5. But I don't need synchronized transcripts
Maybe you already have your interviews transcribed as Word documents - and don't want them synchronized with a media file. This is perfectly fine - just import the transcripts as Document sources and get coding.
6. Plan ahead for auto coding
The structure of a transcript can facilitate auto coding in NVivo - a really useful feature if you want to get your material organized quickly.
For example, if you have a transcript column that identifies each speaker - then you can use auto coding to gather everything a particular person says. Similarly, if you have interview documents with headings that identify each question - then you can auto code based on heading styles and gather all the answers to each question.
The help topic About automatic coding techniques explains all of this in more detail.
7. Ask a researcher
So what is everyone else doing about transcripts?
In her book, Qualitative Data Analysis Practical Strategies, Pat Bazeley has some great insights on the pros and cons of transcribing (check chapter 3).
I'd love to hear what works for you - share your joy (and pain) in the reply box below.