I transcribed my own interview data. I would recommend it. Doing the transcription yourself requires focus and concentration to hear what the participant actually said and transcribe it accurately. I haven't found any software, yet, that is capable of transcribing an audio recording of more than one voice. Most speech recognition programs need to be trained to understand a particular voice to get anything above ~80% accuracy.

However, I recently got a copy of Adobe soundbooth which does speech recognition for several voices. After I played around with it I decided not to use it. Initially it is very inaccurate. I couldn't find any information about how to train it to be better. But if you have time then you might like to investigate. However, I think doing your own transcribing or using a transcription service would be faster and more accurate.

Good luck

For those who are interested I just found a report of a 'software transcription test' for the Adobe Soundbooth CS4 voice recognition software.

http://www.whitehousetapes.net/info/software-transcription-test-adobe-soundbooth-cs4-white-house-tapes

It is a short concise report of the limitations of the current version of the software.

I have used transcription software previously (Naturally Speaking) but I did not find it very reliable and spending much more time correcting myself than doing the transcription myself first hand. In my current research project,
I have recently run four one hour focus groups but because my project has a very short timescale, I had a very small budget for transcription services. I hadn't used a transcription company before and I no-one could recommend one to me. I briefly looked at reviews on the net and also got a few quotes.

Initially, I choose what I thought was a reasonably priced overseas company who had good claims about accuracy (approx $120 per audio hour). I wanted to try initially with just one file as I did not want to waste money if the quality was not very good. I knew the audio files were good quality despite having multiple speakers. I was so disappointed as the transcription was really poor and the sentences weren't coherent at all. I questioned the company (TranscriptionStar) about whether it had actually been a human rather than software and they said it had but I really didn't believe them given the quality.

I then spent a number of hours correcting the transcription but decided it was actually easier to start from scratch. Given the other research activities which needed to be undertaken, I looked once more at different companies and this time decided to phone a couple based in the UK. I spoke to one lady from Transcription Services (http://www.transcript-services.co.uk) who was great. Her name was Carol; she was very friendly and she immediately filled me with confidence given her numerous years of experience and was also very accommodating given my requirements. She offered to do the first 10 mins of transcription for me to test the quality given the bad experience I had previously had. It was great. Such a remarked difference. With the final transcripts I obviously still had to go through and identify the speakers and correct some of the technical terms but I was going to listening to the audios at least one or two times during the analysis anyway. I also need the transcriptions on a very short timescales and for four hour audio files on a 3 day turn around, I was only charged 80p per audio minute. It would have been cheaper if I'd be able to wait a bit longer. Carol was very accommodating, quick to reply to my queries and provided a value for money service.

Sorry if this sounds like a sales pitch, it isn't but I just wish someone had been able to recommend someone to me then I wouldn't have had this initial bad experience. I do think there are some value for money options out there which if you can get a small research budget for this activity, I allows more time on the most important part: the analysis.

I'd recommend outsourcing transcription! I had my transcription (and some translation) done for me in India. My interviews were conducted in India anyway, but I found a great transcription company in Pune that I would highly recommend. They offer a very efficient and reliable service with an ftp server to make everything quick and easy. I was also able to pay in pounds as the company has a UK bank account. I negotiated a rate per hour of recording, and it was very reasonable (less than £10 per hour). They do normally specialise in medical transcriptions, but they are happy to take on any job. http://www.outsourcedmtservices.com/ I liaised with the director Hemant Joshi, you can reach him at hjoshee@eth.net Please quote my name if you get in contact with them. I can't promise how they will be for you, but they were great for me and I would recommend their service highly.

I think outsourcing the trancription is the best method. I also keep the digital recordings so I can always play them back to refresh my memory. I utilise CAT reporters. They use the court machine and transcribe fairly easily. I pay by digital tape recordings. Each qualitative transcription could be more than 20 pages. I found doing them myself is rich yes but quite tedious and time consuming.

Hope this helps