See how your score ranks against the class
A raw mark alone says little about how well you did relative to others. This calculator places your score on the class distribution: it converts your mark into a Z-score, then into a percentile that tells you roughly what fraction of students you outperformed. All it needs is your score, the class mean, and the standard deviation.
How it works
First it computes the Z-score, the number of standard deviations between your score and the mean:
z = (score - mean) / standardDeviationIt then maps that Z-score to a percentile using the cumulative distribution function of the standard normal curve, evaluated with a high-accuracy error-function approximation. A Z of 0 maps to the 50th percentile, a Z of plus one to about the 84th, and a Z of minus one to about the 16th.
Worked example
Suppose you scored 78 on an exam where the class mean was 65 and the standard deviation was 10:
- Z-score = (78 − 65) / 10 = 1.3
- A Z of 1.3 corresponds to approximately the 90th percentile
That means you scored higher than roughly 9 out of 10 students in the class — even though 78% as an absolute mark might seem unremarkable. The raw score and the relative rank are two very different pieces of information.
Now consider a harder exam where the class mean is 55 and the standard deviation is 15. The same score of 78 gives Z = (78 − 55) / 15 = 1.53, pushing you to roughly the 94th percentile despite the identical raw score. Context — meaning the distribution of the whole class — completely changes your standing.
Key Z-score benchmarks
| Z-score | Approximate percentile | Meaning |
|---|---|---|
| +2.0 | ~98th | Top 2% of the class |
| +1.5 | ~93rd | Top 7% |
| +1.0 | ~84th | Top 16% |
| 0.0 | 50th | Exactly at the class mean |
| −1.0 | ~16th | Bottom 16% |
| −2.0 | ~2nd | Bottom 2% |
When is the normal model accurate?
The calculation assumes class scores follow a bell curve. This is a reasonable approximation for:
- Large classes (typically 50+ students)
- Midterms and finals with a wide spread of difficulty
- Standardised tests designed to produce a normal distribution
It is a weaker approximation for very small classes, heavily curved exams where many students cluster around a single score, or assessments with hard score caps that create ceiling effects. In those cases, treat the percentile as an estimate and confirm the actual rank list with your instructor if precision matters.
Where to find the mean and standard deviation
Most learning management systems display class statistics alongside returned grades. If your instructor only posts the mean, you can estimate a rough standard deviation from the range — but this will reduce accuracy. For standardised tests like the GRE or SAT, ETS and College Board publish the exact mean and standard deviation for each administration.