How Grade Curving Works: Adding Points, Scaling, Square Root, and Z-Score Methods
October 21, 2026
When a test turns out harder than expected, instructors sometimes curve the grades to adjust the distribution. The term covers several different methods, which affect students differently. Some give the same bonus to everyone; others help low scorers more than high scorers, or reshape the distribution to a target mean and spread.
Consider a small class with seven scores: 52, 61, 68, 72, 75, 80, and 88. The mean is 70.9. An instructor wants the average to be around 80. Here is how four methods would work.
Skip the math and get your answer instantly:
Open the GPA Calculator →Method 1: Add Flat Points
Add a constant to every score: here, about 9 points to raise the mean to 80. The scores become 61, 70, 77, 81, 84, 89, and 97. It is simple and fair in the sense that everyone gains equally, but a top student could exceed 100, and the gaps between students stay the same. Many instructors cap scores at 100.
This method assumes the test was uniformly too hard.
| Original | Add 9 points | 10 x sqrt(score) | Z-score to mean 80, SD 10 |
|---|---|---|---|
| 52 | 61 | 72 | 63 |
| 61 | 70 | 78 | 71 |
| 68 | 77 | 82 | 77 |
| 72 | 81 | 85 | 81 |
| 75 | 84 | 87 | 84 |
| 80 | 89 | 89 | 88 |
| 88 | 97 | 94 | 95 |
Method 2: Square Root Curve
Take the square root of the score and multiply by 10: 10 × √score. A 64 becomes 80, a 49 becomes 70, and a 100 stays 100. This method gives larger boosts to lower scores: 52 rises by 20 points (to 72), while 88 rises by only 6 (to 94). It compresses the range and is popular when many students struggle. Critics note that it benefits the weakest scores most, which may feel unfair to students who performed well.
Method 3: Linear Scaling
Scale the scores so that the top score becomes 100 (or another target) by multiplying by 100 ÷ highest score. With a top score of 88, multiply by 1.136: 52 becomes 59, 72 becomes 82, and 88 becomes 100. This preserves relative differences proportionally. It only helps if the top score is below 100, and it can leave low scorers still failing.
Method 4: Standardizing (Z-Score) Curve
Convert each score to a z-score using the class mean (70.9) and standard deviation (11.06), then map it to a new mean and standard deviation: new score = 80 + 10 × z. A 52 has z = −1.71, which becomes 62.9; an 88 has z = 1.55, which becomes 95.5. This method sets the class distribution, so grades depend on relative performance. It is used in some large classes, but it can be harsh in strong classes, because the average student is always assigned the same score regardless of absolute performance.
Some instructors curve by assigning fixed percentages of A's, B's, and so on (grading on a curve, strictly speaking).
What It Means for You
Ask your instructor for the policy in the syllabus. A curve that adds points helps everyone equally; a relative curve means your grade depends on your classmates. If a curve is likely, you cannot rely on it, so aim to master the material. And remember that a curve cannot lower your grade in most policies; it is meant to adjust only upward, but check the rules for the course.
Frequently Asked Questions
What does it mean to curve a test?
To adjust scores by a formula so that the distribution matches a desired mean or shape.
Does a curve ever lower grades?
Most instructors use curves only to raise scores, but strict relative grading can lower individual grades.
Which curving method is best for low scorers?
The square root curve gives the largest boosts to the lowest scores.
What is the 'bell curve' grading?
Assigning grades based on position in the class distribution, with fixed percentages for each letter.
Should I count on a curve?
No. Not all instructors curve, and the method can vary.