6.5 Recommended Ways to Use Problem Sets
Where Practice Tests give you the headline diagnosis, Problem Sets are where the actual teaching happens. A practice test tells you a student is weak on Inference questions or Advanced Math; a Problem Set is how you close that specific gap, one focused session at a time.
This article steps back from the mechanics of building a Problem Set and looks at the bigger picture: when to reach for one, how to choose what goes inside it, and how to slot it into a student's week so it actually moves their score. Read this alongside the earlier articles in Section 6, which cover the how-to.
The Core Loop: Test, Diagnose, Drill, Re-Test
The single most important habit in EdisonOS is this loop:
Practice Test → Identify Gap (from the Test Report) → Targeted Problem Set → Re-Test
Practice Tests diagnose. Problem Sets treat.
Reports confirm whether the treatment worked. A tutor who runs this loop consistently for every student will outperform one who simply assigns content in order, every time. Problem Sets are not a warm-up before the "real" test, they are the lesson the test pointed you toward.
The Four Patterns
1. Gap-Closing Problem Sets (Most Common).
Build directly from the Student Response Review tab of a recent practice test. If the report shows a student missed 7 out of 9 Inference questions, build a Problem Set drawn from your Inference question folder and assign only that. Keep these short like 5 to 10 questions is usually enough for a single focused session. The goal is repetition on one pattern until the student can recognise it cleanly, not exposure to a wide mix.
2. Concept-Introduction Problem Sets.
Use these when you are teaching a new topic for the first time, before the student has seen it on any test. Pull questions from an easier folder, keep the set short, and pair it with a tutoring session where you walk through the first few questions together. This is the lowest-pressure way to use a Problem Set it functions almost like guided practice.
3. Mixed-Review Problem Sets.
Once a student has worked through several gap-closing sets, build a Problem Set that mixes question types they've recently studied. This forces them to identify what kind of question they are looking at a skill that disappears when every set is single-topic. Use these every two to three weeks as a checkpoint.
4. Pre-Test Warmup Problem Sets.
A short Problem Set (10–15 questions) drawn from the topics a student is weakest on, assigned the day before a Full-Length practice test or simulation. The goal here is not to teach anything new it is to prime the student so they walk into the test having recently seen their hardest question types.
Choosing What Goes Inside a Problem Set
This is the part most tutors get wrong. A well-built Problem Set follows three rules.
Stay narrow. A Problem Set covering "Reading & Writing" is too broad to teach anything. A Problem Set covering "Inference questions from short literary passages" is sharp enough to build a real skill. The narrower the focus, the faster the student improves on it.
Match the difficulty to the goal. If the student is learning the concept, pull from easier folders. If the student already understands the concept but keeps missing the questions, pull from harder folders the gap is in execution, not knowledge, and easier questions won't expose it. Use the same Question Bank folders you use for Practice Test sections; the difficulty tagging carries over.
Keep it shorter than you think. 10 to 20 questions per set is the sweet spot for focused practice. A 50-question set turns into homework the student grinds through without thinking. Two 15-question sets with a tutoring discussion in between will outperform one 50-question set every time.
Problem Sets vs Practice Tests: When to Reach for Which
Problem Set | Practice Test | |
|---|---|---|
Primary purpose | Teach and reinforce one specific skill | Measure overall performance |
Length | Short typically 10–20 questions | Long - full module or full-length |
Scope | Narrow - one question type or topic | Broad - full subject or full exam |
Timing | Usually untimed or relaxed | Timed, often with strict rules |
Frequency | Multiple per week | Once a week or once every few weeks |
Built from | A specific gap surfaced by a Practice Test | Diagnostic need, exam prep cycle |
Student mindset | "I'm learning this" | "I'm being tested on this" |
The simplest rule: a Practice Test tells you what to work on; a Problem Set is how you work on it.
A Sample Week for One Student
To make this concrete, here is what the loop looks like in practice for a student two months out from their exam.
Monday: Review last week's Full-Length practice test together. Identify two weak areas — say, Inference (R&W) and Linear Equations (Math).
Tuesday: Student completes a 15-question Inference Problem Set. Tutor reviews wrong answers in the next session.
Wednesday: Student completes a 15-question Linear Equations Problem Set.
Thursday: A Mixed-Review Problem Set drawing from the two topics plus one earlier weak area, to check whether the gains are holding.
Friday: Pre-test warmup Problem Set - 10 questions across this week's focus areas.
Saturday: A Reading & Writing Module Test to measure whether Inference has actually improved in test conditions.
Sunday: Off, or light review.
Notice how Practice Tests bookend the week and Problem Sets do the daily teaching work in between. This is the rhythm the platform is built for.
Habits That Make the Difference
Build Problem Sets from real reports, not intuition. Open the student's most recent Test Report's Student Response Review tab before you build the next Problem Set. Let the data choose the topic. A tutor who "feels" the student is weak on geometry is sometimes right and sometimes wrong; the report is always right.
Don't skip the debrief. A Problem Set without a follow-up conversation is just homework. The teaching happens when you walk through why the student got the wrong ones wrong, the Problem Set itself is just the raw material for that conversation.
Reuse and rotate. Keep a small library of trusted Problem Sets organised by topic. When you need to re-test whether a gap has actually closed, assign a fresh Problem Set on the same topic rather than the same questions the student has already seen.
Watch the trend, not the single result. One Problem Set's score is noise. A student going from 50% to 70% to 85% on three Inference Problem Sets over two weeks is a signal and a signal you can show the student, which is powerful motivation in the final weeks before an exam.
Where to Go Next
Practice Tests and Problem Sets together form the core teaching loop on EdisonOS. The patterns you build here are what those reports will measure.