A quiz score tells you a student got 6 out of 10. It doesn't tell you whether the four they missed were all the same skill, scattered across four different skills, or all a careless mistake on the same type of question. That distinction is the difference between a five-minute reteach targeted at one misconception and a full class period re-covering material most students already have.
Reading the pattern, not just the total
Item analysis just means tallying which specific questions got missed, and by how many students, instead of only recording the total score. You don't need software for this on a ten-question quiz — a tally sheet with one row per question and a checkmark per wrong answer takes five minutes.
Once you have the tally, three patterns show up:
1. Everyone missed the same question — that's a teaching gap, not a student gap. Reteach the whole class.
2. A handful of students missed the same cluster of questions — that's a small-group reteach, not a whole-class one.
3. Wrong answers are scattered with no pattern — that's more likely to be a testing-condition issue (rushed, distracted) than a content gap, and reteaching the content won't fix it.
Turning the tally into a plan
Once you know which pattern you're looking at, AI is useful for the next step: designing the actual reteach, not just naming the gap. Try:
- "Here's what a class of 24 missed on a quiz: [list the specific questions and the misconception each one reveals]. Design a 10-minute warm-up activity that directly targets [the specific misconception], different from how I originally taught it."
Asking for a *different* explanation than your original one matters — if the first explanation didn't land for a chunk of students, repeating it verbatim usually doesn't either.
A quick before/after
Before: a quiz average of 74%, recorded in the gradebook, with a vague plan to "review before the test."
After: a tally showing 9 of 24 students missed the same two questions, both testing the same misconception about negative exponents, with a 10-minute targeted warm-up planned for Monday using a different explanation than the original lesson.
Where this gets easier to build into your routine
Our [quiz generator](/quiz-generator) can tag each question by the specific skill or misconception it's testing when you build it, so the item analysis step is already half-done by the time students turn quizzes in. Our [differentiation planning tool](/differentiation-plan) is useful for the small-group version of the reteach — building a targeted activity for the handful of students who missed the same cluster without slowing down the rest of the class.
What still needs your judgment
- A single quiz is one data point. Don't restructure a unit off one rough quiz if the pattern doesn't repeat on the next check.
- Distinguish a content gap from a question-wording problem. Sometimes "everyone missed it" means the question was ambiguous, not that the class doesn't understand the material — reread the question itself before assuming it's a teaching gap.
- Some misconceptions need more than 10 minutes. A targeted warm-up handles a narrow gap; a genuinely foundational misunderstanding may need a full reteach lesson, and item analysis is what tells you which one you're facing.
