Diagnostic assessment: what to check before you teach anything

The class in front of you in September is not the class that left in June. Some students spent the summer reading, some forgot half of what they knew, a few came from a different school with a different syllabus, and all of them are now mixed into new groups. Teaching the first unit as if everyone starts from the same place is a guess — usually a wrong one.

A diagnostic assessment replaces that guess with data. This guide explains what a diagnostic assessment is, how it differs from formative and summative assessment, why the first week of school is the moment to run one, six concrete examples you can use tomorrow, and a step-by-step 10-question diagnostic you can build in an afternoon (or generate in seconds).

What is a diagnostic assessment?

A diagnostic assessment is an assessment given before a unit, a course or a school year begins, to establish a baseline: what students already know, which prerequisite skills are solid, and which misconceptions they carry in. The word diagnostic is borrowed from medicine on purpose. A doctor doesn't prescribe before examining; a diagnostic assessment is the examination that comes before the teaching.

Three features define it. It is prior to instruction — you run it before you teach the content, not after. It is low-stakes — no grade, no consequence for the student, which is exactly what makes the answers honest. And it is actionable — the point is not the score but what you do with it: which topics to skip, which to reteach, and which students need something different from the rest.

You will also hear it called a pre-assessment, a pre-test or a baseline assessment. In everyday classroom use these are the same thing; if there is a nuance, "pre-test" usually means a quiz that mirrors the end-of-unit test, while "diagnostic" is broader and includes tools like self-ratings and concept checks.

Diagnostic vs formative vs summative assessment

Teachers are often told there are three (or four) types of assessment, and the names blur. The clean way to separate them is by when they happen and what the result is for.

The fourth type sometimes added is ipsative assessment — comparing a student to their own earlier performance rather than to a standard. A diagnostic assessment at the start of the year is, conveniently, the baseline that makes ipsative comparison possible later.

If you want the formative column in depth — the tools, the routines, what to look for — we covered it in 10 formative assessment tools reviewed. Diagnostic is what comes before all of that.

Why the first week of school is the moment

Any moment before a unit is a fine time for a diagnostic. The first week of the school year is the best one, for three reasons that don't apply at any other point.

The class has changed over the summer. Skills that were automatic in June are rusty in September, unevenly so: the same lesson that would have been a review for everyone in spring now lands on students who are months apart. A diagnostic tells you how far apart.

The class is new. Groups are reshuffled, students arrive from other schools or programmes, and you have no record of what last year's teacher actually got through. What the previous curriculum says was covered and what the students can do are two different things; only one of them can be planned from.

You haven't committed yet. In the first week you can still reorder the first unit, add a two-day review of a prerequisite, or split the class for a topic. By week four the plan has momentum and every adjustment costs more. The diagnostic is cheapest exactly when it is most useful.

There is a fourth reason, less about planning: the first-week diagnostic sets a tone. Run it as a game, tell students honestly that it doesn't count and that wrong answers are useful, and you have shown them in the first hour that this classroom is a place where not knowing yet is normal.

Six diagnostic assessment examples

Diagnostic assessments come in more shapes than "a test at the start". These six cover most classrooms and most subjects; you can mix two or three in the same week.

1. Prerequisite pre-test. A short multiple-choice quiz — eight to twelve questions — on the skills the first unit assumes rather than the ones it teaches. Fractions before algebra, sentence types before essay writing, the periodic table before bonding. The results tell you whether the foundation is there. This is the workhorse and the one we build step by step below.

2. Concept inventory. A quiz where every wrong answer is a specific wrong idea. The classic is the Force Concept Inventory in physics, where the distractors are the intuitive beliefs students bring in ("heavier objects fall faster"). You don't need a validated instrument: write questions where each distractor encodes a misconception you meet every year, and the answer distribution tells you which ones to attack first. Our guide to writing multiple-choice questions explains how to build distractors that do this job.

3. K-W-L chart. Three columns: what I Know, what I Want to know, what I Learned. Students fill the first two on day one; the third waits for the end of the unit. It is qualitative, fast, works from age seven upwards, and it surfaces both prior knowledge and curiosity — which the multiple-choice pre-test doesn't.

4. Confidence self-rating. Give students the unit's learning objectives in plain language and ask them to rate each: I can do this / I've seen it but I'm not sure / New to me. Then compare the ratings with the pre-test. The students who are confident and wrong are the ones the pre-test alone would have hidden.

5. Muddiest point from last year. One open question: "What is the one thing from last year's course you never really understood?" Anonymous, two minutes. You will get the same three answers from a third of the class, and they are your first review lesson.

6. Two-minute conferences. While the rest of the class works on the pre-test or the K-W-L, talk to each student for two minutes with one or two questions from the pre-test asked aloud. It is the slowest method and the richest: you hear how they think, not just whether they got it.

A 10-question first-week diagnostic, step by step

Here is the pre-test recipe we recommend. Ten questions, about fifteen minutes of class time, and a result you can act on the same afternoon.

Step 1 — Pick the three prerequisites. Look at your first unit and list the skills it assumes from last year. Choose the three that, if missing, would break the unit. Write one question each (questions 1–3). Keep them at the level students should have reached, not harder.

Step 2 — Pick the four core concepts. Take the first module you will actually teach and choose its four key ideas. Write one question each (questions 4–7) at the level you would ask after teaching them. Most students should get these wrong — that is the point. The ones who get them right are the students the module will bore.

Step 3 — Plant two misconceptions. Think of the two wrong ideas that come up every single year in this topic. Write one question each (questions 8–9) where the misconception is one of the answer options, and make it the most tempting one. The share of students who pick it is your most useful number in the whole diagnostic.

Step 4 — End with one honest question. Question 10 is open and about the student, not the content: "Which of these topics do you feel least ready for, and why?" or "What's one thing you want me to know about how you learn?" It is the only question you will read one answer at a time — and the one students remember you asked.

Step 5 — Run it as a low-stakes live round. Tell the class it doesn't count, that you expect them to miss some, and that this is you finding out how to teach them. Run it live so you see the answers as they come in, or assign it as homework if you'd rather they take their time.

Step 6 — Shortcut: generate it. If you have last year's syllabus, a chapter PDF or even a list of topics, you don't have to write the ten questions by hand. PanQuiz's AI Quiz Generator builds a multiple-choice set from a topic description in seconds, and the free PDF to Quiz tool does the same from a document — upload the chapter you will teach first and you have questions 4–7 ready to edit. Prefer paper for day one? The printable quiz maker turns the same set into a worksheet with an answer key.

How to read the results (and what not to do with them)

A diagnostic assessment is only as good as the reading you give it. Three rules.

Read per question before per student. The first thing to look at is the answer distribution of each question across the whole class. Question 2 at 90% correct means prerequisite two is fine — skip the review. Question 8 with 60% of the class on the misconception distractor means you have found your first lesson. Per-question results are what a class snapshot looks like; a list of student scores is not.

Then read per student, in groups. Sort students by pattern, not by score. The students who missed the prerequisites need a different first week from the ones who aced the core concepts, and both are different from the student who was confident and wrong on the misconception questions. Three or four groups is usually enough; a live quiz platform gives you the per-student, per-question grid to do this in a few minutes — the same view a classroom response system gives you during a lesson, applied once, at the start.

Don't grade it, don't rank it, don't post it. The moment a diagnostic assessment affects a grade, students stop answering honestly and start guessing strategically — and the data is worthless. Keep the results between you and each student. If you want to use it later, the honest use is ipsative: give the same ten questions at the end of the unit and show each student how far they moved.

Run this way, the first-week diagnostic costs one lesson and pays for itself by week two: the review you didn't need to teach, the misconception you caught before it fossilised, and the four students you found on day two instead of at the first test.