Guide8 min read

How to compare two assessments fairly

Why a rise from 60% to 66% may mean nothing, how to compare topic by topic, and a calculator that separates real change from noise.

Who it's for
Teachers, tutors and institute leads who report improvement.
What it helps with
Avoid over-claiming or missing real progress when comparing tests.
Written by
Heramb.study editorial team
Last reviewed

"Average went up from 60% to 66%" sounds like progress. It might be. It might also be an easier paper, a different mix of topics, or normal variation. Before reporting improvement — to a student, a parent or your management — check three things.

1. Are the tests comparable?

Compare like with like: the same topics, similar question types, similar length and time limit. A follow-up check that only includes the topics you practised is fine, as long as you compare it against the same topics in the first test — not against the whole first paper.

2. Compare topic by topic, not totals

A stable total can hide a big gain in one topic and a loss in another. Topic-level comparison shows both, and tells you where your teaching made a difference.

3. Allow a margin for noise

Two different tests are never exactly equal, and students have good and bad days. Treat small changes — a few percentage points — as no clear change. The calculator lets you choose the margin. Use a wider margin when topics have only a handful of questions.

Assessment comparison calculator

Runs in your browser — marks you type, paste or upload aren't sent to us

Enter each topic's percentage on two assessments. Changes within the margin are reported as "no clear change".

5 is a sensible default when the two tests are similar; use 8–10 if they differ in length or difficulty.

What to do with the result

  • Improved: move the topic to maintenance — occasional mixed practice.
  • No clear change: check whether the practice targeted the actual cause of mistakes.
  • Declined: first check the questions were comparable; then look at the mistake types.

Record results in a progress report so the student can see the change too.

Sources

  1. Black, P. & Wiliam, D. (1998). Assessment and classroom learning. Assessment in Education: Principles, Policy & Practice, 5(1), 7–74. doi.org/10.1080/0969595980050102

Spotted an error? Report it — we correct mistakes and update the review date.

Invite-only early access

Help shape Heramb.study from the first version.

We're opening access in small, invite-only groups so we can learn from real classrooms before scaling. Requesting an invite is free and doesn't commit you to anything.