Transcribe research interviews
Upload your fieldwork recordings and get timestamped transcripts with speaker labels that you can correct, export and bring into your coding workflow.
How it works
- Upload each recording and note which participant code it belongs to.
- Select the language the participant speaks and start the transcription.
- Listen back, correct the text and check speaker labels against your notes.
- Export DOCX or TXT and load it into your analysis software.
Qualitative research lives and dies on the transcript. A dissertation chapter, a user study or a policy evaluation may rest on forty hours of recorded conversation, and turning those hours into text is a tax on the time you would rather spend on analysis. Scribewren produces a timestamped draft for each recording so your hours go to reading and coding instead of typing.
A draft is the right word. Before a transcript enters your dataset, you listen along and correct it, paying attention to names, technical vocabulary and passages where the participant was quiet or two people spoke at once. Do this the same way for every file. A consistent approach to fillers, pauses and unclear speech matters more to your analysis than shaving a few minutes from one transcript.
Consent and handling
Handling of recordings is a research decision, not just a technical one. Your ethics approval may specify where audio can go and how long it can be kept. Scribewren deletes originals within 24 hours of processing and does not use your files for training, but you are the one who decides whether that fits your protocol.
Tips
- Name files with a participant code, not a real name, so transcripts match your data management plan.
- Decide on verbatim or clean style before you start, and apply the same editing rules to every transcript.
- Keep your original recordings in your own approved storage; the originals are removed from Scribewren within 24 hours.
- Check your ethics approval for the terms under which recordings may be processed by a third party.
Questions
Is the transcript verbatim?
It follows what was said, including repetition where the speech was clear. Cleaning up filler words and false starts is an editing choice you make afterwards.
Can I get timestamps for coding?
Yes. Every line has a timestamp, and exports keep them if you choose a format that supports it, such as SRT or VTT.
Is participant data used to train models?
No. Your files and transcripts are never used for training, and the original upload is deleted within 24 hours of processing.
Can I transcribe participants who speak other languages?
Scribewren supports 38+ languages, including many that are commonly used in field research. Check that yours is on the list and test with a short clip first.
How many recordings can I upload together?
Unlimited accepts up to 50 files in one upload. Free takes one at a time.
Try it on your own recording
Free includes a few transcriptions a day. No card needed. Compare plans.
Start free