How Researchers Can Anonymize Participant Video for Presentations

Before using participant video in a research presentation, check whether the study's consent and approval cover that form of sharing. Permission to record a session does not, by itself, tell you whether a clip may appear at a conference, in a paper supplement, or on a public website.

Choose the smallest excerpt that supports the research point.

Separate the observation from the identity

Ask what viewers need to see to understand the finding. A hand movement, interaction with an interface, or task sequence may be sufficient without a recognizable face.

If facial expression is the object of study, face treatment can remove useful information. That conflict needs a methodological and ethics decision rather than a casual editing choice.

A recreated example or descriptive figure may sometimes communicate the point more appropriately.

Match the output to the approved audience

An internal lab meeting and a recorded public talk have different exposure. Check whether the conference records sessions or whether the slides will be distributed.

Keep the participant's original recording under the study's access rules. Make a derivative for the approved presentation and record the source interval.

For regulated or sensitive research, follow the institution's tool and vendor requirements before any processing.

Treat faces and other identifiers separately

Where appropriate and authorized, use selective face treatment on the working clip. Inspect repeated appearances, profiles, and small video-call tiles.

Then check voice, screen names, task data, account details, and the interview content. A participant can be identifiable from an unusual experience or a distinctive environment after their face is hidden.

Unseen currently focuses on faces, so plan separate edits for text, screens, or audio. Its local mode changes the media-processing location; it does not establish research approval.

Preserve the validity of the example

Do not cut away hesitation or rearrange events in a way that changes the finding. If masking hides a relevant interaction, tell the audience or choose another presentation format.

Keep a note of transformations so another authorized researcher can understand the derivative.

A face-treated clip is not automatically an anonymized dataset. Use the study's defined terminology and assessment.

Check the slides and recordings

A title slide, screenshot, speaker notes, or filename can reveal details omitted from the video. Create stills from the reviewed derivative.

If the presentation will be recorded, review the package for that wider audience before delivery.

For product usability sessions, the user-research guide gives concrete examples of replacing real account data and shortening highlight clips.