Categorizing Test Mistakes by Root Cause

The most common mistake when reviewing incorrect answers is simply copying down the correct solution and moving on. An error log made that way gives you a false sense of understanding ("Oh, of course!"), only for you to repeat the exact same mistake on the next exam. What you need to record is not the answer key, but why you fell into that specific trap. Every paragraph in this prompt is designed to isolate that thinking process.

This is why the Task paragraph strictly confines the causes to four predefined categories. If you leave the categories open-ended, AI tends to default to vague phrases like "insufficient conceptual understanding" for nearly every error. When calculation slips and conceptual gaps get lumped together, your study strategy becomes ineffective. Providing an explicit "Undetermined" option prevents forced misclassifications. If you only provided a one-line answer with zero scratch work, it is natural that the cause cannot be pinpointed—and that serves as a reminder to show your work next time.

Breaking down the evaluation into three Steps forces the AI to reverse its default analysis order. Left unprompted, an AI immediately outputs the standard textbook solution, meaning your actual thought process never appears on screen. By demanding "Inferred Thought Process Behind My Answer" first, the focus shifts from copying answer keys to diagnosing cognitive patterns.

In the Format paragraph, the table structure is built to serve directly as an actionable error log. The summary tally below the table is especially useful: if six out of ten errors stem from calculation slips, you need better double-checking habits, not more theory lectures. Finally, wrapping the input in """ delimiters prevents the AI from mistaking problem text (like "Which of the following is true?") as system instructions directed at itself.

Asking this way produces long textbook explanations for each problem. Since that information is already in your study guide, you gain little from reading it again. It completely fails to reveal whether multiple mistakes share a common root cause, leaving you to figure out what to prioritize on your own.