Grading Curve

Curving a Class’s Test Scores

A grading curve is a class-wide adjustment applied to every student’s raw score, usually to raise scores after a test turned out harder than intended. Enter every student’s raw score (on the standard 0-100 scale), pick a curving method, and this calculator shows each student’s adjusted score plus the new class average.

This is different from a single student’s own grade calculation — it looks at the whole class’s scores at once, rather than one student’s weighted average or how many questions they can afford to miss.

Choosing a Curve Method

• Flat (add a fixed number of points) is the simplest method: every student’s score goes up by the same fixed amount, e.g. “add 5 points to everyone.”
• Square root curve takes the square root of each raw score and multiplies it by 10. It boosts lower scores by a much bigger margin than higher ones, since it’s mathematically self-limiting — it can never push a score above 100.
• Curve to a target average finds the flat number of points needed to raise the class’s current average up to whatever target you choose, then applies that same amount to everyone.

Every method here only ever adds points (or adds zero) — never subtracts — and every adjusted score is capped at 100, since a curve should never push a result above the maximum possible score.

The Formulas

The flat curve simply adds a chosen number of points, $p$, to every raw score:

$$\text{Adjusted Score} = \min(100,\ \vA{\text{Raw Score}} + \vB{p})$$

The square root curve takes the square root of the raw score and scales it back up to the 0-100 range:

$$\text{Adjusted Score} = \min\left(100,\ \sqrt{\vA{\text{Raw Score}}} \times 10\right)$$

The curve-to-target-average method first finds the shift needed to bring the class average up to the target, $t$ (never a negative shift — a class that already meets or beats the target gets no adjustment):

$$\text{Shift} = \max\left(0,\ \vB{t} - \vA{\text{Class Average}}\right)$$

and then applies that same shift to every student, capped at 100 exactly as the flat curve is.

Worked Examples

1. Flat curve, +5 points: raw scores of 60, 70, 80 become 65, 75, 85 — the class average moves from 70 to 75. 2. Square root curve: a raw 64 becomes \sqrt{64}{' '} \times 10 = 80 (a 16-point boost), while a raw 90 becomes $\sqrt{90} \times 10 \approx 94.9$ (under a 5-point boost) — the lower score gets the bigger lift. 3. Curve to a target average of 75: raw scores of 50, 60, 70 average 60, so every score gets a 15-point shift, becoming 65, 75, 85 — a new average of exactly 75.

Key Factors to Consider

• Different curve methods send different messages about a difficult test. A flat curve treats every point of the test as equally “off,” while a square root curve implicitly treats a low score as more likely due to test difficulty than a high score — choosing a method is partly a judgment call about why the class performed the way it did, not just arithmetic.
• Curving doesn’t change the underlying rank order among students (for a flat or curve-to-target shift). Adding the same number of points to every score preserves who scored higher than whom — only the square root method can shift relative standing somewhat, since it compresses the range between high scores while expanding the range among lower ones.
• A curve is a policy decision, and many institutions have specific rules about when and how it’s allowed. Some schools or departments have explicit policies restricting or standardizing how instructors may curve grades — always check institutional policy before applying a curve to official grades.
• This calculator assumes a standard 0-100 percentage scale for every score. If your grading scale is different (e.g. points out of a non-100 total), convert scores to a 0-100 percentage basis before entering them here for these formulas to apply correctly.

Common Mistakes

• Entering raw points out of a non-100 total instead of converting to a percentage first. As covered above, every formula here assumes a 0-100 scale — a score of “42 out of 50” needs to become 84 before it’s entered, not left as the raw point value.
• Assuming a curve preserves rank order regardless of method. As covered above, a flat curve and curve-to-target both preserve who outscored whom, but the square root curve genuinely changes relative standing by compressing the gap between high scorers and widening it among low scorers — don’t assume the method choice is purely cosmetic.
• Applying a curve without checking institutional policy first. As covered above, curving is a grading-policy decision, not just a math operation — some schools restrict or standardize how and when instructors may curve, and skipping that check can create a real compliance problem even if the arithmetic itself is correct.
• Choosing “curve to a target average” as if it guarantees every student passes. The shift raises the average to the target, but it’s still a flat addition — a student who scored far below the rest of the class can still end up below the target average after the same shift is applied to everyone.

Useful to Know

• The three methods encode different assumptions about why the test came out low: flat curve assumes the whole test was equally too hard, square root curve assumes weaker performers were hit hardest by difficulty (a common assumption when a test has a few very hard questions), and curve-to-target simply targets a specific class average without a stance on why scores were low.
• Because every method here caps at 100, a curve can never inflate a strong score past a perfect one — only the students below the ceiling actually see their score move.
• For a single student’s own “what do I need” question rather than a whole-class adjustment, the Grade Calculator and Test Score Calculator answer that from the individual side, while this calculator works from the class-wide side.