What are authenticity checks?

At Prolific, we're committed to maintaining the highest standards of research quality. To support this goal, we've introduced authenticity checks - a powerful tool that helps researchers ensure they're receiving genuine responses from participants.

Authenticity checks are technology-enabled verification processes that help researchers identify when free-text responses are generated using AI tools (like ChatGPT) or copied from external sources rather than written by participants themselves.

These checks are designed to protect the integrity of research on our platform by distinguishing between genuine participant responses and those created through automated means or other methods of non-original content generation. Most importantly, they ensure that honest participants like you can continue to participate in quality research.

How do authenticity checks work?

Authenticity checks analyze behavioral patterns that may indicate when participants are using third-party sources to answer free-text questions. The system looks for certain indicators without reading the content of your responses.

If our system detects patterns consistent with AI-generated text or copied content in a free-text response, the submission may be flagged for the researcher to review.

Good news for participants: Our authenticity verification system has been carefully calibrated to achieve 98.7% accuracy, with an extremely low false-positive rate of just 0.6%. This means that participants providing genuine responses have very little to worry about.

Will authenticity checks read my responses?

No. Authenticity checks don't examine what you've written. They only identify behavioral patterns that suggest the use of third-party sources to generate answers. Your privacy remains a top priority.

When are authenticity checks used?

Like attention checks, authenticity checks are an optional tool that researchers can choose to implement in their studies. They only work for free-text questions, so not every study will use them.

Currently, authenticity checks are available for studies run on:

Qualtrics
Gorilla
Prolific's AI Task Builder

When can researchers use authenticity checks?

Researchers are instructed to use authenticity checks responsibly. If a study legitimately requires you to research or use external sources, researchers should not implement authenticity checks for those specific questions.

Importantly, researchers cannot reject your submission based on authenticity checks if their study explicitly requires you to use external sources or tools.

What should I do if I believe I've been falsely flagged?

While our system is highly accurate (98.7%), we understand that no technology is perfect. If you believe your submission was incorrectly flagged by the authenticity check system with the rejection reason ‘Failed authenticity check’:

Contact the researcher directly through Prolific's messaging system as your first step
Explain your situation clearly and respectfully
If the issue remains unresolved after speaking with the researcher, please contact our support team who can look into this for you

We take false positives very seriously and will investigate any cases brought to our attention.

This is something that we want to evolve and adapt so that this is always getting better at doing what we want it to do - so feedback is crucial

Tips for success with authenticity checks

To ensure a smooth experience with studies that may include authenticity checks:

Read study instructions carefully - they should indicate when you should or shouldn't use external sources
Provide your own thoughts and opinions - remember that your authentic perspective is what researchers value most!
If you're uncomfortable with a study's requirements, you can always return it without penalty

Our commitment to fair participation

At Prolific, we're dedicated to creating an environment where truthful participants can thrive. Authenticity checks help us maintain research quality while ensuring that honest participants like you can continue to make valuable contributions to research.

By implementing these checks, we're working to create a more trustworthy platform where genuine responses are valued and rewarded.

Need further help?
Click here to contact us