A Diagnostic Quiz
When I get it wrong, say which step I got wrong
Build a diagnostic quiz from the material below: [paste material]. Requirements: 1) first list the 4–6 steps involved in understanding this topic, e.g. identify the variables, pick the right formula, handle units, interpret the result; 2) one or two questions per step — few, but aimed at that step; 3) after I answer, diagnose which step I failed before explaining why that step is easy to get wrong; 4) if I miss the same step twice in a row, stop and re-teach that step before moving on.
How to use
The step list is the whole trick — without it, the model can only tell you the answer was wrong.
Good for
- You've done plenty of practice problems, your accuracy is stuck, and you can't tell which link is weak
- You're coaching a junior or tutoring a student and need to know whether the block is the concept, the mechanics, or the reading of the question
Not for
- The goal is passing an exam with a fixed format; past papers plus error review match the real difficulty better
- The material is purely factual (dates, names) with no reasoning steps to decompose
Example
Input
Quadratic word problems. The student sets up the equation correctly every time, but the final answer is slightly off.
What you get
It should decompose into set the variable, write the relation, solve, then check units and domain, and locate the failure at sign handling in step three or at dropping the negative root in step four. Then it writes two questions aimed only at that step, and on a miss it names the sign error when moving terms instead of saying careless.
Common mistakes
- Steps so coarse that the diagnosis collapses back into 'the basics are weak', which suggests nothing you can act on
- Question difficulty spanning too wide a range, so a miss can't be attributed to the step or to the question being out of scope
Try tweaking it like this
- Flip it: 'I'll explain each step first, then you write questions that test it', which works better for self-study
- Add 'score my misses by step across three attempts', to separate a systematic gap from a random slip