Interpreting Quiz Analytics: Beyond Completion Rates
Completion rates tell you whether a quiz works. Question-level analytics tell you why. Here is how I read quiz data to find the specific problems worth fixing. For two years I looked at one metric to evaluate my quizzes: completion rate. If it was high, the quiz was good. If it was low, the quiz was bad. This told me whether a quiz was working but never why. When a quiz had a fifty percent completion rate, I had no idea which question was causing users to abandon. I was flying blind. Once I started reading question-level analytics, I could diagnose specific problems and make targeted fixes. Here is how I interpret quiz analytics now. Question-Level Drop-Off The single most diagnostic metric is question-level drop-off, which is the percentage of users who abandon the quiz at each question. Most quizzes have a smooth, gradual decline where a small percentage of users leave at each question. A spike in drop-off at a specific question is a red flag. Something about that question is pushing users out. I investigate every drop-off spike by reading the question carefully. The usual culprits are: the question is too hard, the question is too long, the question is ambiguous, or the question is boring. More guides are on the Quiz Hub Blog.