About this tool
A free, browser-based grade curve calculator. Paste a class set of raw test scores, pick one of four curve methods, and see the curved results side by side with the originals. The statistics card shows count, minimum, maximum, mean, median, and standard deviation for both the raw and curved distributions, so you can compare the effect of a curve before you commit to it.
Everything runs in your browser. No accounts, no tracking, no analytics. The scores you paste never leave the page, and they disappear when you close the tab.
The four curve methods
Each method has a different shape and a different defense. Try them all on the same data — the worked examples below use the built-in sample, which has a mean of 73.36 and a max of 98.
Flat add
Adds a fixed number of points to every score. With +5 on the sample, every student's score moves up by exactly 5: the 38 becomes 43, the 73 becomes 78, the 98 becomes 100 (clamped). This is the easiest curve to explain to students and parents because no one's relative position changes. Use it when a test ran a few points harder than usual across the board, and you want a clean, transparent adjustment.
Square root curve
Replaces each raw score with the square root of the score, times ten. The classic pattern: 49 becomes 70, 64 becomes 80, 81 becomes 90, 100 stays 100. On the sample data, the 38 becomes 61.6, the 73 becomes 85.4, and the 98 becomes 99.0. Low scores get the biggest lift; the top is barely touched. Use it when the test was clearly too hard and you want to compress the bottom of the distribution upward without ballooning the top above 100.
Linear scale to max
Scales every score by the same factor so the highest score becomes a target value (default 100). If the sample's max is 98 and the target is 100, every score is multiplied by 100/98 ≈ 1.0204: the 38 becomes 38.78, the 73 becomes 74.49, the 98 becomes 100. Use it when the highest performer in your class fell short of the ceiling but you trust the test was otherwise fair — the relative rank order and proportional gaps stay intact, but the top is anchored at your chosen max.
Set average to target
Shifts every score by the same amount so the class average lands on a target value (default 75). On the sample, the raw mean is 73.36; targeting 75 adds 1.64 to every score. Unlike the flat add, the shift can be negative — if your raw mean is already above target, every score moves down. Use it when you have a year-over-year benchmark for what the class average should be on this material, and you want to align this distribution with that baseline.
When to curve, when not to
A curve is a correction for a measurement problem — the test didn't capture what students actually learned. The classic cases are an unintentionally hard question, an ambiguous item that students reasonably misinterpreted, or a topic that turned out to need more class time than you'd budgeted. In those cases, a curve restores the signal the test was supposed to carry.
What a curve isn't good for: hitting a target letter-grade distribution every quarter, papering over a class that genuinely didn't engage with the material, or rescuing a grading scheme that needs structural rethinking. If every test gets curved, the curve stops being a correction and becomes the actual grading scale — at which point you should rewrite the rubric instead.
One practical check: look at the BEFORE/AFTER standard deviation. If a curve dramatically changes the spread of the class, you're not just shifting grades — you're changing what the test measures. Flat add and set-average-to-target preserve standard deviation; square root and linear-scale-to-max change it.
Frequently asked questions
When should I curve grades?
When the score distribution doesn't reflect what students learned — usually because a test was harder than intended or an item was ambiguous. Don't curve to hit a target distribution every time; that hides real signal.
What is the square root curve?
New score = √raw × 10. A 49 becomes 70, a 64 becomes 80, a 100 stays 100. Lifts low and middle scores the most while leaving the top alone.
Can a curve lower a student's score?
Yes — 'Set average to target' shifts every score by the same signed amount, which is negative when the class is already above target. 'Clamp at 100' caps above 100 but doesn't prevent downward shifts. Check the AFTER column before applying.
How do I pick a curve method?
Flat add is the easiest to explain. Square root helps when scores are low across the whole class. Linear scale anchors the top performer at the ceiling. Set average to target aligns with a historical benchmark. Try each and compare the AFTER statistics.
Is my data private?
Yes. The calculator runs entirely in your browser. Nothing is sent, stored, or logged. Close the tab and the scores are gone.
Can I undo a curve?
The raw scores stay in the textarea exactly as you typed them. Switching methods or clearing a parameter restores the original column instantly — the BEFORE statistics never change.