A test-prep session structure that actually works

Vikrant Singh·July 3, 2026·7 min read

Vikrant Singh is co-founder and Chief Operating Officer of Slide Practice. He writes about running a one-to-one practice.

Diagnostic-driven, error-log based, timed under real conditions. The session arc I run with SAT and ACT students, plus how to structure the final weeks.

A test-prep session that works follows the same arc every time: a short warm-up on last week's errors, one skill taught to the point of use, a timed practice set under real conditions, a full review of every miss, and homework tied to the next session's diagnostic focus. The structure stays fixed so the student's attention goes to the material and not to guessing what comes next. Sixty minutes, five parts, driven by an error log instead of a page count.

Why a fixed structure beats a fresh plan every week

Most test-prep sessions drift because they are organised around content instead of errors. You open the book to where you left off, work through some questions, and stop when the hour is up. The student leaves with a vague sense that they did "some math," and next week you do it again a few pages later.

The trouble is that page count is not the thing you are trying to move. The score moves when specific, repeated mistakes stop happening. So the spine of your prep should be an error log, not a textbook. Every wrong answer, and every right answer the student was unsure about, goes into a running list with the reason it happened. That list tells you what to teach next. It also tells the student that prep is finite: there are only so many kinds of mistake they make, and each session closes a few of them.

A fixed structure supports this. When the shape of every session is the same, you stop spending the first ten minutes deciding what to do. The student knows a timed set is coming, so they stop treating practice as low-stakes. And you get a clean before-and-after every week: did last session's fix hold, or did the error come back? If you want the wider picture of building a tutoring practice around this kind of repeatable system, I keep more of it on the page for tutors.

The five parts of a session

Here is the arc I run for a standard sixty-minute session. Stretch the minutes for a ninety-minute block, but keep the order.

1. Warm-up review (5 to 10 minutes)

Start with two or three questions the student missed last time, pulled straight from the error log. Do not re-teach yet. Just have them try again cold. If the fix held, you both see it in thirty seconds and you move on. If it did not, you have found the first real target of the day. This opening also settles the student and signals that nothing gets logged and then forgotten.

2. Targeted skill (10 to 15 minutes)

Teach one thing. Not four. The single skill comes from the largest cluster in the error log: comma splices, systems of equations, inference questions, ACT Science graph reading, pacing on the Reading and Writing module. Name it out loud, teach the rule or the process, then work two examples together where you think aloud and they watch, then hand it over. Resist the urge to cover everything you noticed. One skill taught to the point where they can use it beats five that got mentioned.

3. Timed practice set (15 to 20 minutes)

This is the part most tutors skip or soften, and it is the part that matters most. Put a visible timer on the table. Give a set of questions at real test pacing, roughly the seconds-per-question of the actual exam. No stopping, no hints, no "wait, let me explain." The student marks anything they are unsure of and keeps moving. You are training two things at once: the skill you just taught, and the ability to perform it under time pressure, which is a separate skill entirely. A student who solves a problem with unlimited time and misses it in ninety seconds does not have a content problem, they have a timing problem, and you only see that under the clock.

The digital SAT gives about 71 seconds per Reading and Writing question and closer to 95 on Math; the ACT is faster, near 36 seconds on English and 60 on Math. Match your set to the pace the student actually has to hit, not to how long the problem takes without a clock.

4. Error review (10 to 15 minutes)

Now go through the set. Review every miss, and also every question they got right but flagged as a guess. For each one, sort the mistake into a type: content gap, careless slip, misread the question, ran out of time, or second-guessed a right answer. These categories matter because the fix is different for each. A content gap needs teaching. A careless slip needs a process change. A misread needs a habit, like underlining what the question is actually asking. Have the student say the correct path out loud rather than nod at your explanation. Then log it.

The most useful column in an error log is not the right answer. It is the one-line reason the student got it wrong, in their own words.

5. Homework and next diagnostic focus (5 minutes)

Assign a short, specific set on today's skill, plus one old error-log question to re-attempt cold next week. Write down what next session's diagnostic focus will be, usually the second-largest error cluster. Then send a short recap so the work does not evaporate between sessions. Homework that connects directly to what you just did in the room gets done far more often than a generic chapter assignment. If getting work done between sessions is your real struggle, that deserves its own approach, and I wrote one out in getting students to do homework.

The session plan template

Copy this, fill it in before each session, and keep the filled-in copies. Over a few weeks they become a record of exactly how the student changed.

TEST-PREP SESSION PLAN
Student: ______ Date: ______ Target test: SAT / ACT
Test date: ______ Sessions remaining: ______

1. WARM-UP REVIEW (5-10 min)
- Re-do 2-3 questions missed last session (from the error log)
- Did last week's fix hold? Yes / No
- Quick recall: [rule, formula, or strategy from last session]

2. TARGETED SKILL (10-15 min)
- Today's ONE skill: ______________________
- Why today (from the error log): ________________
- Teach the rule, then 2 worked examples together, then hand over

3. TIMED PRACTICE SET (15-20 min)
- ____ questions in ____ minutes on a visible timer
- Real test pacing: about ____ seconds per question
- No stopping, no help. Mark any question you're unsure of.

4. ERROR REVIEW (10-15 min)
- Review every miss AND every "right but guessed"
- Tag each: content gap / careless / misread / timing / second-guess
- Log it: question, topic, why it was missed, the fix
- Student explains the correct path out loud

5. HOMEWORK + NEXT DIAGNOSTIC FOCUS (5 min)
- Assign: ____ practice set on today's skill (due next session)
- Plus 1 error-log question to re-attempt cold
- Next session's diagnostic focus: ________________
 (the biggest remaining error cluster)
- Send recap: what we did, the fix, the homework

Structuring the final weeks

The five-part session is your engine for most of the prep window. As the test date closes in, the balance shifts from building single skills to rehearsing the whole test.

Six weeks out and earlier: build skills

Run the structure as written, one skill per session, working down the error clusters from largest to smallest. The error log is short and specific at this stage because you are mostly discovering what the student does not yet know.

Three to five weeks out: full-length practice tests

Add a full-length timed test every week or two, taken in one sitting under real conditions: same time limits, same breaks, phone away. The test itself becomes your diagnostic. The two or three sessions after it are pure error review, cluster by cluster, feeding straight back into the log. This is also when timing problems surface that never showed up in short sets, because stamina and pacing only break down over a full section.

The final week: taper, do not cram

Page count against error log, side by side

A few mistakes to avoid

• Teaching more than one skill. A packed session feels productive and retains almost nothing. Pick the biggest cluster and stop there.
• Skipping the timer. Untimed practice hides the exact weakness the test will expose. Use the clock from the first week.
• Reviewing only the misses. A right answer the student guessed is a future miss. Flag and review those too.
• Letting the log live in your head. If it is not written down, next week you will teach from memory and repeat yourself.
• Assigning generic homework. "Do chapter seven" gets ignored. A short set tied to today's skill gets done.

Keep the record between sessions

Two things carry a student between one session and the next: the error log and the recap. The log is your teaching plan. The recap is the student's memory of what changed and what to practise. Both should be quick to produce, or you will quietly stop doing them by week four. A recap can be five plain lines, always the same shape. If you want a format for the notes themselves, I laid one out in how to write tutoring session notes, and you can see what a finished recap looks like in this sample recap.

None of this needs software. A shared doc, a timer, and an error log you actually keep will carry a student all the way to test day. If you would rather have the recap and the log write themselves from what happened in the session, so you spend the time teaching instead of typing, that is the small thing Slide does for independent tutors, and it takes no commission, ever. You can see a real one on the recaps page.

Common questions