It occurred to me after finishing my post on Friday (where I showed what transcripts completed by Otter.ai looked like using a few app options) that I also should compare my performance to Otter, so I transcribed the same passage. I use the title "Old school," because I did this with a word processing program, no specialized software, and using the play/pause/rewind controls on the website recording of this interview with Walter Adams.
This took me 9 minutes. It is very accurate, I believe but not impressively fast. If I had used a pedal and software that allowed me to slow the recording (I like playing things somewhere in the 92-96% of real time range), I probably could have done this a little faster. I listened three times total (included in the time) and the last time basically was just a final check. To get a really good idea of the time cost for this process, I need to compare myself going from scratch to someone who uses Otter to build the basic framework. While that would be interesting, it is a challenge to rationalize use of duplicate effort just out of curiosity. Maybe a better alternative is to expand this into a class assignment and have students compare their own performance on a small segment of tape, as I just did. There is, of course, the increased intimacy with data gained by transcribing from scratch - and that is pretty difficult to quantify.
27 Comments
I just downloaded the Otter app (free version; limited to 600 mins of recording) for iOS and did a quick trial using open access archived audio interview. The cost for Otter premium right now is $8.33 per month and it provides 6000 minutes.
The interview recording I used was of Walter C. Adams reflecting on the May 4, 1970 Kent State University shootings. The quality was nowhere near usable, but, on the other hand, it was surprisingly accurate in spots. It was interesting to watch it self-correct as the speaker continued. If you listen to the recording (click on the link above), you will notice that it is not of the highest quality, there is some noise through the whole recording. On the other hand, he speaks very clearly. I do not know yet whether having this framework as a start for a transcript will be beneficial, harmful, or make no noticeable difference when compared to transcribing from scratch myself. For those people who grew up thumb typing, a program like this is probably going to be helpful. For an old, I mean experienced typist/keyboardist like myself - and when I describe myself as a "keyboardist," I refer to computers and musical instruments - any time savings via this program is likely to b highly variable depending on the recording and the speaker(s). I also personally like the sensation of typing-while-listening, maybe it reminds me of doing music dictation (writing-note-while-hearing-music) many years ago... I recorded just over one minute, and the processing time is pretty slow. It took about 18 mins to generate these few lines. It was a lot like waiting while exporting a screencast video (I use Screenflow) from the program format to Mp4 - basically you should just find something else to do instead of watching the minutes tick away. Pluses - it looked for and tagged keywords that you can use to generate a search. Maybe this would be beneficial to start first cycle coding. Otter also records - so for a live interview is an interesting option (although using a tablet/phone is not necessarily secure). Minuses - free version will only export to plain text, but maybe this is not really so bad. If you click on readmore, I have screenshots of some of the alternative ways to save as text files. These can obviously be pasted into word processing, so that in itself is not a good reason to upgrade. And 600 minutes per month may go pretty far for some people... |
AuthorI am Sheryl L. Chatfield, Ph.D, C.T.R.S. I am a member of the faculty in the College of Public Health at Kent State University. I also Co-coordinate the Graduate Certificate in Qualitative Research and I am a member of the Design Innovation Team at Kent State. Archives
February 2024
Categories
|