How to Plan a Review Week Before a Big Test That Actually Raises Scores

You know the trap. The unit test is Friday, kids look nervous, and you cave and spend Monday reteaching Chapter 1 like it's August. By Thursday you've covered everything again, nobody has actually practiced anything, and the scores look about the same as last year. Review week only raises scores when students do the thinking, not you. Here is how I run it.

Start with the data, not the calendar

Before you plan a single day, pull the last two quizzes and any exit tickets. Sort the questions by standard or topic and count wrong answers. You will almost always find that 70 percent of the misses cluster in three or four ideas. For a stoichiometry test that might be limiting reagent, mole to mole ratios, and unit tracking. For a kinematics test it is usually sign conventions, choosing the right equation, and interpreting graphs.

Plan the week around those clusters. The stuff most kids already have does not need a day. It needs five minutes of retrieval and a move on.

Retrieval first, reteach only if it fails

Every day should open with a low stakes retrieval task, closed notes, five to eight minutes. Not a warm up puzzle. Actual test-shaped questions from the weak clusters. Then you grade it live, by show of hands or a quick walk around, and here is the important part: only reteach the specific step where they broke.

If 22 out of 28 kids set up the ratio correctly but botched the unit conversion, do not reteach limiting reagent. Do a two minute board example on the conversion and move to more practice. Reteaching what most kids already know is the number one time sink in review week.

Mix the problems on purpose

Blocked practice, twenty limiting reagent problems in a row, feels productive and transfers badly. Interleaved practice is uglier in the moment and works better on the test. On day three or four, hand out a set where problem types are shuffled and students have to decide which approach to use before they can solve. That decision is the actual skill the test measures.

A simple version: pull eight problems from the last three quizzes, scramble the order, and require students to label the type before solving. It takes ten minutes to build and it exposes the exact confusion that tanks test scores.

Make error analysis the homework

Instead of assigning more problems, hand back their old quizzes and have them do a written error analysis on three missed items. I use three columns:

• What I did
• Where it went wrong (be specific, not 'careless')
• What I will do next time, in one sentence

Kids hate it the first time and then it becomes the most useful thing they do all week. You can grade a class set in fifteen minutes because you are only reading the third column.

A sample five day shape

• Monday: diagnostic retrieval quiz on the whole unit, then error analysis on the two weakest topics
• Tuesday: targeted mini lesson on cluster one, followed by fifteen problems, mixed difficulty
• Wednesday: mini lesson on cluster two, interleaved practice set, partner check
• Thursday: full length practice test under real conditions, no notes, timed
• Friday morning: go over the practice test, kids annotate their own errors, then the real test if you like, or push it to Monday

Thursday is the one most teachers skip and it is the one that moves scores the most. A timed practice test tells you and them exactly where they still are.

The honest part

Building a week like this from scratch takes a few hours you probably do not have, especially if you teach three preps. That is the whole reason we built Rested Teacher, to hand you a full year of lessons and review sequences already structured this way so you can spend your prep time on the kids, not the plan.

Whatever you use, the shift is the same. Less talking from you, more retrieval and mixed practice from them, and a hard look at the data before you decide what to reteach. That is the difference between review week as a comfort blanket and review week as something that actually shows up on the scores.