Did you know that two students can both improve on a test yet have very different stories behind their scores? Looking only at the final mark hides where each learner began, what the test covered and whether the comparison is fair.
To compare before-and-after scores use three checks: record each starting score, calculate the change in percentage points when both tests use the same percentage scale, and confirm that the assessments are sufficiently comparable. Then inspect which skills improved on a fresh task. A larger score increase is useful evidence to investigate, not automatic proof of greater effort or better teaching.
Diagnose the comparison before doing arithmetic
Ask first: are both results from the same subject, at a comparable level of difficulty, with similar timing, support and marking? A 65% score on difficult unfamiliar questions cannot be compared straightforwardly with 85% on rehearsed questions and treated as pure learning growth.
Next, decide what you are trying to learn. Are you measuring one child’s progress across time, comparing achievement levels, or evaluating whether a specific practice routine helped? Those are three different questions. Before-and-after scores can contribute evidence but do not by themselves prove why a change happened.
Worked example: two learners, two different starting points
Imagine two students taking comparable practice assessments, each marked out of 100. Their scores are invented to explain the calculation; they are not eduKate student results or claims about tuition outcomes.
- Student A starts at 40/100 and later scores 60/100.
- Student B starts at 75/100 and later scores 85/100.
- A’s change is 60 − 40 = 20 percentage points.
- B’s change is 85 − 75 = 10 percentage points.
Student B has the higher final score: 85 versus 60. Student A has the larger observed change on the reported scale: 20 percentage points versus 10. Neither statement contradicts the other. Achievement and change are different quantities.
A percentage-point change is not the same as a percentage increase. For Student A, the relative increase in score is (60 − 40) ÷ 40 × 100% = 50%. For Student B, it is (85 − 75) ÷ 75 × 100% ≈ 13.3%. Saying that A’s score “rose by 50 percentage points” would be incorrect. In everyday school reports, percentage points are often the clearer way to describe a change in test percentage.
There is another caution. Student B had only 25 percentage points of headroom below 100 at the start; Student A had 60. A smaller change near the ceiling does not necessarily mean weaker learning. Percentages are bounded and may not behave like equally precise measurement scales across the full range.
What to check before concluding that learning improved
1. Compare like with like
Check topic coverage, total marks, allotted time, allowed aids and marking standards. If the second paper is easier, the numerical improvement may overstate the change in competence. If the second paper is harder, it may hide real gains.
2. Separate practice familiarity from transfer
Repeating the identical ten questions can improve recall of their answers rather than the underlying skill. Use a new but suitably comparable question to check whether the learner can apply the idea independently.
3. Read the error patterns, not just the total
A change from seven correct to nine correct answers might occur because the learner repaired a specific misconception, slowed down to check arithmetic, or simply encountered easier items. Look for the underlying reason in the workings and the types of questions missed.
4. Notice uncertainty
Short tests can be noisy. One rushed day, a misunderstood instruction or a small number of items can change the percentage considerably. Treat a single comparison as a clue and look for the same skill on later independent attempts.
A practical parent–student score conversation
Use a short record: date, topic, result, conditions, first recurring mistake and next check. Ask “Which questions can you now solve that you couldn’t solve before?” rather than “Why didn’t you gain as many marks as your friend?” The former invites evidence of learning; the latter confuses progress with competition.
For a learner who started high, look for deeper independent explanations, fewer repeated errors and success on slightly less familiar problems. For a learner rebuilding basics, monitor stability with the prerequisite skill before celebrating a single higher total.
How tutors and teachers can use this comparison
- Keep the score scale consistent and identify changes in test conditions.
- Compare topic-by-topic evidence where the tests cover different mixes of skills.
- Choose one representative follow-up question without a model answer visible.
- Distinguish improvements in accuracy, understanding and speed rather than collapsing them into one number.
- Tell families what the scores support and what remains uncertain.
A tutor should not turn two practice scores into a promised grade jump. They are snapshots of performance, and the best next instructional decision usually comes from the actual work shown, not only from the headline score.
Mistakes that make progress look larger or smaller than it is
- Calling a rise from 40% to 60% a “20% increase” without saying which definition you mean.
- Comparing results from papers with different difficulty as if they were equivalent.
- Treating the learner who starts higher as if they had more room to improve.
- Mistaking a familiar repeated question for independent transfer.
- Ignoring whether the child had hints, extra time or the answer key during the second attempt.
- Assuming that a higher score proves a particular intervention caused the change.
For an independent checking step, give the learner one new question targeting a previously incorrect skill. Ask them to solve it unaided and explain why the method fits. Record the date and the degree of help, if any. Repeat after some time. This yields more useful diagnostic evidence than retallying the same totals.
When score comparisons are limited
A numerical score is only one lens on education. Speaking, collaboration, creativity, persistence and the quality of explanation can be important without being captured by a short paper. Assessment also has context: illness, access needs and changing school tasks can affect performance. Do not pretend one percentage measures a whole child.
If the results come from different grading systems or different maximum scores, normalise cautiously and check the test content. Merely converting both marks to percentages does not make two unrelated assessments equivalent.
Frequently asked questions
Does the student with the bigger score increase learn more?
Not necessarily. A larger observed increase on comparable tests is useful evidence of changed performance, but starting point, difficulty and measurement uncertainty matter.
Should families compare siblings or classmates this way?
It is usually more helpful to follow each student’s own skill development. Different starting points and learning needs make simple rankings misleading.
Can a score decrease even when the child learned something?
Yes. A harder test, different topic mix, stress or a small sample of questions can lower the observed score. Inspect the questions and conditions before deciding what changed.
Make the next comparison more informative
Keep the baseline, use a comparable follow-up and add one independent skill check. That turns “the mark went up” into the more useful question: “What can the learner now understand, explain and do alone?”
Continue through the Education Hub, the related Core Aim discussion of practice papers with different difficulty, and the Clementi Mathematics guide on diagnosing the exact weakness and verifying independent application.