Interview consent form

The consent form governs everything you may subsequently do with your interview data. What it does not say, you may not later do. This page provides a template to adapt and explains, for each clause, why it is there and what breaks when it is missing.

The mistake that happens most often

The most widely circulated template names purpose, voluntariness and anonymisation. It does not name who technically processes the recording. Anyone who uses it and then runs the interviews through a transcription tool has obtained no consent for that step.

This rarely surfaces until somebody asks, and by then it cannot be repaired, because collection is finished. The order therefore matters: decide what will transcribe, then draft the consent, then collect.

Template

To copy and adapt. Replace the bracketed placeholders; the superscript numbers point to the explanations below.

Why each clause is there

1. Purpose limitation

The GDPR requires the purpose to be specified at collection. An open formulation such as “for research purposes” does not hold, because it would cover any later use. The more concrete the purpose, the more robust the consent. If you want to reuse the material in a follow-on project, you need a separate basis for that.

2. The processor belongs in the form

This is the clause missing from most templates. As soon as an external service processes the recording, that is processing on your behalf under Article 28 GDPR. You need a data processing agreement with the vendor and informed consent from the interviewee.

If you transcribe locally, meaning entirely on your own device, both fall away. In that case say so, because it is an advantage worth stating.

The full picture is in Interview transcription and the GDPR.

3. Describe anonymisation honestly

Promise only what you can deliver. Someone who guarantees “fully anonymised” and then leaves the organisation name in the transcript has breached their own undertaking. With expert interviews genuine anonymisation is particularly hard, because position and sector often identify the person. In those cases it is cleaner to promise pseudonymisation and agree the published version separately.

4. Two periods, not one

Recording and transcript have different lifespans. The recording is the most sensitive artefact and goes first; the anonymised transcript may stay longer. A single shared period forces you either to delete the transcript too early or keep the recording too long.

5. Withdrawal without disadvantage

Voluntariness is a validity condition. In interviews inside your own organisation, or with students in a dependent relationship, that is not self-evident, and it is exactly where scrutiny lands. The explicit statement that no disadvantage arises belongs in the form.

6. Data subject rights

The list is short and standard, but its absence is noticed. If your institution has a data protection officer, add their contact details.

Check before the first interview

1. The transcription tool is chosen before the form is drafted
2. A data processing agreement is in place, or processing runs locally
3. The processing location appears in the form
4. The anonymisation promise matches what will actually be done
5. Separate periods for recording and transcript are entered
6. The signed form exists before recording starts
7. The participant keeps a copy

Note: this template is a practical aid, not legal advice. Check whether your institution has its own mandatory wording before adapting it.

Once the recordings exist

After collection comes the part that costs the most time. Nodl takes the existing audio file, separates the speakers where there are several, and gives every passage a timestamp that jumps straight to that point in the audio.

Two points matter for what you write in the form: processing and storage take place in Germany, the language models run inside the EU, and content is not used to train models. A single recording may be up to one hour long.

Common questions