University of Bergen · Evidence judgement study

Participant information and privacy details

TEST-ONLY DRAFT · This copy is for technical review. Do not recruit participants or treat this page as final participant-facing approval.

Please read this page before deciding whether to take part. The study start page contains the consent checkboxes. This page explains the study, what participation involves, what information is used, and how to ask questions or raise a data-protection concern.

1. Who is organising the study?

The study is called Disclosed Assistant Reliability and Delayed Evidence Judgement. It is conducted by the Department of Information Science and Media Studies at the University of Bergen (UiB), Norway. The project manager and scientific lead is Christoph Trattner. The study team’s contact address is shown at the bottom of this page.

2. What is the purpose of the study?

The study examines whether practising with a scripted assistant that is described as highly reliable or as occasionally mistaken affects how people later judge evidence without assistance. It is research about evidence judgement and learning, not an assessment of your health, intelligence, or personal suitability for any treatment.

3. Who can take part?

The planned participants are adult UK residents aged 18 or older who can complete the study in English and are recruited through Prolific. The protocol requires desktop or laptop participation. Do not take part if you are under 18.

4. What will I be asked to do?

Participation is planned across three online visits:

The rewards are £3.00 for Visit 1, £1.50 for Visit 2 and £2.40 for Visit 3, paid separately for each completed visit. Payment does not depend on the accuracy of your answers or whether you complete later visits. If you voluntarily withdraw partway through a visit, you will be compensated for the time spent on that visit at £9 per hour, up to that visit’s full reward. If a technical problem on the study team’s side prevents you from completing a visit you started, you will receive the full reward for that visit. Payments for completed visits will not be withheld because of low accuracy, disagreement with the assistant, non-return for a later visit, or honest reporting that you used outside help. The study team will review Prolific submissions and aims to approve them within 1–2 working days.

5. What materials will I see?

You will read short summaries of real research and assertions written by the study team for this experiment. An assertion is not necessarily a quotation from, or an endorsed conclusion of, the original researchers. Some assertions deliberately go beyond or conflict with the evidence supplied in the packet. You will be asked to judge the exact assertion using only that packet:

“Withhold” does not simply mean that a claim sounds exaggerated or that every imperfect claim should be withheld, while insufficient evidence does not mean that the claim should be rejected. Please do not use external websites or other AI tools while answering. The materials may concern health-related research, but they are for research purposes only and do not provide personal medical advice or recommendations.

6. What is the assistant and what might be misleading?

The assistant is scripted for this experiment; it is not a live AI service. In one practice condition, its suggestions match the study’s adjudicated answer key on all 15 practice items. In the other, 12 of 15 suggestions match that key and three are incorrect. The condition is disclosed as part of the study. Both groups receive explanations of the evidence after practice items. A separate assistant is used in the final additional block and may be wrong. These descriptions concern agreement with the study’s answer key, not a claim that real-world AI or medical advice is infallible.

7. What information will be collected?

The platform is designed to record information needed to run the visits, analyse responses and administer payment, including:

The study does not ask you to provide your name or medical history. The research identifier is pseudonymous, not anonymous: the operational database links it to the Prolific ID so visits and payment can be managed. Do not include identifying details in any free-text response if one is presented.

8. Why is the information processed, and what is the legal basis?

The purpose of data collection is scientific research. The legal basis for processing personal data for research and education purposes is that processing is necessary for the performance of a task carried out in the public interest. Information is used to administer participation, connect visits, make payments, assess study outcomes, check study operation and report research findings. Results will be reported in a form that does not identify individual participants.

9. Who is responsible for the information, and where is it stored?

The University of Bergen (UiB) is responsible for the study and acts as the data controller. The technical deployment is done on NREC – Norwegian Research and Education Cloud as the hosting environment. Access to identifiable operational data is restricted to authorised study personnel who need it for participant administration, payment, data quality or approved research tasks. The data may be temporarily stored and processed on devices owned by UiB.

10. How long will information be kept?

The project is scheduled to end on 12 June 2027. Personal information used to administer participation—including your Prolific ID and the link between that ID and your research identifier—will be deleted by that date. The research responses will be retained in anonymised form for future research and reproducibility, and the anonymised research dataset may be made publicly available. Before public sharing, the dataset will exclude Prolific IDs, linkage keys, IP addresses and information that could identify participants, including identifying combinations of demographic details, timestamps or free-text responses. The research team will assess the dataset to ensure that it is anonymised before release.

11. Is taking part voluntary? Can I withdraw?

Taking part is voluntary. You may stop participating, close the page or choose not to return for a later visit. If you voluntarily withdraw partway through a visit, you will be compensated for the time spent on that visit at £9 per hour, up to that visit’s full reward. If a technical problem on the study team’s side prevents you from completing a visit you started, you will receive the full reward for that visit. Payment does not depend on accuracy. If you withdraw during the study, the platform may still make explanations for material already shown available through your personal link.

If you want to withdraw and ask for your data to be deleted, contact the study team through Prolific before 12 June 2027, providing enough information for the team to identify your study record. Until personal information and the identifying link are deleted, the team may be able to locate and remove your individual record, subject to applicable research and legal requirements. By 12 June 2027, the identifying information and link will be deleted and the research data retained in anonymised form. After irreversible anonymisation, it will no longer be possible to identify and remove your individual contribution. Closing the browser does not automatically delete data already submitted.

12. What are my data-protection rights?

Depending on the applicable law and circumstances, you may have rights to access your personal data, request correction or deletion, restrict or object to certain processing, and complain to the relevant supervisory authority. Some rights can be limited where the applicable research provisions allow this. To exercise a right or ask how it applies, contact the study team using the details below. These rights can only be applied to information that can still be linked to you; after the identifying information and link are deleted by 12 June 2027 and the research data are anonymised, the team will no longer be able to locate your individual contribution.
You can read about your privacy rights in the University of Bergen's privacy policy. For questions about personal data, you can contact UiB's Data Protection Officer, Janecke Helene Veim, at personvernombud@uib.no.

13. Are there risks or direct benefits?

No direct personal benefit is promised. The task involves reading and evaluating research claims; some health-related statements may be inaccurate or contradicted by their accompanying evidence. This may cause mild confusion or discomfort. The study is not a substitute for professional advice. Explanations and a debrief are provided to clarify the intended interpretation of the materials.

14. How will findings be shared?

Findings may be presented in academic publications, reports or presentations. An anonymised research dataset may also be made publicly available for future research and reproducibility. The dataset will exclude direct identifiers and information that could identify you indirectly. The research team will not use your responses to make decisions about your healthcare or eligibility for services.

15. Questions, concerns or withdrawal requests

For questions about the study, payment, withdrawal or a request concerning your data, contact the study team via the contact address at the bottom of this page and, where relevant, through Prolific. You may also contact the University of Bergen’s Data Protection Officer, Janecke Helene Veim, at personvernombud@uib.no. You can complain to the Norwegian Data Protection Authority (Datatilsynet).

Consent

The consent checkboxes are on the study start page. Please return there after reading this information and decide whether you wish to participate. Selecting the consent checkboxes confirms that you are at least 18 and agree to take part; it does not remove your right to stop participating later.