Quality checks and fraud prevention come from multiple layers for Research Flow studies.
Before launching your research study, you'll also want to make sure that you're not getting fraudulent responses and that you have qualified participants responding to your research study.
The in-built mechanisms of User Interviews and Prolific, together with Typeform’s mechanisms of additional custom screening, time-to-complete, and even video and voice formats help prevent copy and paste scenarios.
Typeform
Typeform’s built-in mechanisms are the following:
- Screener questions: the screener task of a research study can filter out unqualified respondents before they reach research questions.
- Session status pipeline: partial sessions aren’t counted as complete responses until the respondent completes the entire research study.
- AI moderation: dynamic follow‑up questions make submitting bot responses harder than static surveys.
Our panel providers (Prolific and User Interviews) also have their own built-in fraud check methods. These features can be used for verifying participants and preventing fraudulent responses.
Prolific
Prolific prevents fraud and verifies participants using a multi-layered framework called Protocol.
- This system combines bank-grade ID checks via Onfido (an Entrust company).
- It requires live video selfies.
- It performs over 110 automated pre-screening and IP deduplication checks.
- It performs surprise re-verifications.
- It includes real-time in-study bot and AI detection tools.
Read on to find out more about Prolific’s layered approach to data integrity.
User Interviews
User Interviews prevents research fraud and verifies participants by combining a multi-layered automated risk model.
- This model analyzes over 50 signals.
- It performs strict digital identity checks during signup.
- It provides ongoing profile re-verification.
- It includes mandatory researcher feedback after every completed session.
Read on to find out more about User Interviews’ fraud deterrence policies.
The ability to flag fraudulent responses in research studies from the Sessions table of your Results is not supported yet.