The number of questions varies depending on your assessment settings and the other validation signals available. There's no fixed number of questions students need to answer correctly, and they don't need perfect answers to demonstrate their understanding.
What determines how many questions students see?
Two key factors influence the number of questions:
Your assessment settings: the AI Use Level and Validity Threshold you select (Essential, Comprehensive or Definitive). Assessments that allow more AI use rely more on the Q&A.
The student's other validation signals: including Human Factors and Authorship Continuity, informed by their writing sample and how well their submitted work aligns with it.
Students who already demonstrate strong evidence through their writing and authorship continuity may see only a few questions. Where more evidence is needed, additional questions give students further opportunities to demonstrate their understanding.
About the questions
Questions are based on the student's own submission, rather than general subject knowledge or new material. Typos and informal language are fine - the focus is on understanding, not spelling or perfect phrasing.
The Q&A takes about 10-15 minutes for most students. We recommend they start at least 15 minutes before the deadline.
If a student isn't validated, they can return to drafting and try again before the deadline, with new questions each time, or send their work to you anyway.
“Not validated” describes the work, never the student. As an educator, you can review the attempt the student chose to send, including their questions and responses, and see how the different validation factors contributed to the result. This provides a starting point for further discussion, alongside your institution's usual academic integrity procedures and professional judgement.