Participant Information Sheet

A. Purpose of the Study

This study is about how people review short videos and decide whether they are real or AI-generated. We want to test whether an experimental human-AI review aid helps people inspect selected parts of a video, make more accurate final real-vs-AI judgements, and explain what they noticed. Participants review each video in isolation and decide whether it is AI-generated or real, as well as a score how real the video is.

B. Why You Are Invited

You are invited because you are an adult aged 18 or older. You should have normal or corrected-to-normal vision and be able to complete the online video-review task on a device that can play short videos. You do not need any technical background.

C. What Participation Involves

The online session will take about 30 minutes and includes:

1. Introduction and consent: You will read the study information, confirm that you understand what participation involves, and answer basic eligibility questions.
2. Video review task: You will watch short videos of physical scenes, some real and some AI-generated. For each video, you will watch the video, complete the assigned review procedure, then judge whether the video seems real or AI-generated and rate how real it seemed on a 1-5 scale. If your session includes introductory feedback, the app will tell you whether each of the first twelve clips was real or AI-generated after you save your judgement. Answers remain hidden for the rest of the task.
3. Review aid task: For some videos, an experimental review aid may select parts of the video for you to look at again. After the assigned second viewing or review-aid step, you may be asked to give your judgement, your realness rating, and a rating of how helpful the second viewing or aid was.

You can complete the study remotely or in person. You should complete it in a setting where you can watch the videos carefully. You may stop participating at any time.

D. Data Collected

The following data may be recorded:

• Eligibility and administration data: Participant ID, study completion status, eligibility answers, and any email address or payment information you choose to provide for payment or prize-draw administration.
• Behavioural responses and human feedback: final real-vs-AI judgements, final realness ratings, helpfulness ratings, strategy comments, and typed feedback.
• Task interaction data: video identifiers, review-aid condition, selected video parts shown for re-review, exhibit opens and viewing time, playback controls and replay behaviour, the playback state at each decision, page focus, response times, task timestamps, completion status, viewport information, and technical loading or error events needed to run, reproduce, and debug the online study.
• Questionnaire data: age, gender, vision status, and other brief demographic or experience questions relevant to the study.

Participants may request deletion of their data using the participant ID by emailing the organisers at kcs41@cam.ac.uk. Pseudonymised study data will be deleted after 5 years. Prize-draw contact details will be deleted immediately after prize verification.

E. Possible Risks or Discomforts

The study involves watching videos and using a computer screen. Some people may experience mild eye strain or fatigue, and repeated judgements or ratings may feel tiring. You may pause if the task instructions allow, and you may withdraw at any time. We do not expect any risks beyond ordinary online computer use.

F. Benefits

There is no direct personal benefit beyond any payment or prize draw described in the study listing or task information. Your participation will help researchers understand whether human-AI review aids can support people when inspecting potentially AI-generated videos, and how such aids should be designed responsibly.

G. Data Privacy, Security, and Storage

• Anonymisation and pseudonymisation: Your Participant ID is used instead of your name. Prize-draw contact details are stored separately from survey responses, can be accessed only by authorised researchers, and are not included in research analysis.
• Storage: Data will be stored on secure, access-controlled systems and handled in accordance with the GDPR and University Research Data Management Policy.
• Retention: Pseudonymised study data will be retained for up to 5 years for analysis and potential anonymised reuse. Prize-draw contact details will be deleted immediately after prize verification.
• Use: Results may appear in academic papers, presentations, or open datasets, but no individual participant will be identifiable.
• Access: Only the named researchers and authorised collaborators involved in the study will have access to raw study data.

H. Voluntary Participation and Withdrawal

Your participation is voluntary. You may stop at any time. If you withdraw before completing the study, payment or prize-draw eligibility and withdrawal handling will follow the study instructions. You may request deletion of your identifiable data where this remains possible. Once data have been anonymised and included in aggregate analysis, it may no longer be possible to remove individual records.

I. Compensation

In Person: You will receive £15 for taking part, even if you withdraw early.

Online: Spot the Fake participants may optionally enter a prize draw for one £100 Amazon gift card after completing all cases. Any other payment will follow the study listing and task information. Read the raffle guidelines.

J. Ethical Approval and Contacts

This study will be reviewed by the Department of Engineering Ethics Committee, University of Cambridge, before recruitment begins. If you have questions about the study, contact the researchers listed above. For independent advice about your rights as a participant, contact: Department of Engineering Ethics Committee (research-ethics@eng.cam.ac.uk)

K. Summary of Your Rights

• Your participation is voluntary.
• You may withdraw at any time.
• Your data will be pseudonymised or anonymised where possible.
• No identifiable information will be published.
• You can contact the research team with questions or concerns.

L. Optional Referral Challenge

After completing every case, you may optionally compete to be the campaign's highest referrer. The participant with the most valid completed referrals receives one £50 Amazon voucher. You receive a three-word link and only friends who visit that link before starting and then complete every assigned case count. Joining or declining does not affect your study participation or the separate £100 prize draw.

The referral system stores a campaign-specific pseudonym, your three-word code, completed referral credits, and a contact email used only to administer the reward. It stores no raw study user ID, handle, assignment, judgment, response, or raffle key. The pseudonym can be recomputed only by an authorised operator holding the separate secret, so the data are pseudonymised rather than fully anonymous.

Referral contact details are deleted after prize verification. Participant-level referral profiles and credits are deleted after the 30-day dispute window. Clearing your study cookie or changing devices means the app cannot recover your referral status. Read the referral terms.