From raw marks to a grade
A test score becomes a grade in two steps that are worth keeping separate. The first is arithmetic: divide the points earned by the points possible and multiply by 100. That step has one right answer. The second is policy: decide which letter that percentage represents. That step has no universal answer at all — it is set by a syllabus, a department, or a district, and the boundaries move.
Nearly every dispute about a grade is really a dispute about the second step. The 90/80/70/60 cut-offs used here are the most common American convention, and the plus/minus refinement that puts an A at 93 and an A− at 90 is close behind, but plenty of institutions use a seven-point scale where an A starts at 93 and a B at 85, and plenty of individual instructors set their own. This calculator makes the arithmetic exact and shows you the bands it applied so you can substitute your own.
The second output worth attention is the value of a single question. On a 50-point test each mark is worth 2 percentage points; on a 20-question quiz each item is worth 5. That number tells you how much a single scoring dispute is worth and how volatile the assessment is. Short assessments produce coarse, jumpy grades, which is a reason to weight them lightly rather than a reason to avoid them.
The formula and the two ways to feed it
The percentage is E ÷ M × 100, where E is points earned and M is points possible. Nothing else belongs in it. If partial credit was awarded, E is not an integer, and that is fine.
When every item is worth the same, you can skip counting the credits and count the errors instead: E = M − w, so the percentage becomes (M − w) ÷ M × 100. That is what teachers mean by an "easy grader" — a table of percentages indexed by number wrong for a test of a given length. The chart this calculator prints is exactly that table, generated for whatever length you enter.
Two adjustments change the answer and both are policy choices. Rounding decides borderline papers: 86.67% is a B under plus/minus bands if left alone and a B+ if rounded to the nearest whole percent, because rounding lifts it to 87. Pick the rule before you grade the stack, apply it to every paper, and state it on the syllabus. The calculator flags any case where the rounding rule changes the letter.
The second adjustment is bonus items. If a test carries extra-credit questions that are not in the denominator, a student can exceed 100%. The arithmetic handles that without complaint — nothing in the formula caps the ratio at one — but decide in advance whether your gradebook caps it, because a 104% on one test can quietly offset a failure elsewhere.
Finally, note that the letter and the grade points are not the same currency. The letter is what appears on the paper; the grade points are what that letter is worth if the score becomes a course grade in the GPA calculator. A single test rarely becomes a course grade directly — it enters through the syllabus weights instead.
Worked example: 43 out of 50, and the same paper three other ways
A student scores 43 points on a 50-point test.
- Divide. 43 ÷ 50 = 0.86.
- Convert to a percentage. 0.86 × 100 = 86.00%.
- Find the value of a point. 100 ÷ 50 = 2.00% per point, so the 7 points lost cost 14 percentage points.
- Assign the letter. On plus/minus bands, 86 falls between the B floor of 83 and the B+ floor of 87, so it is a B, worth 3.0 grade points.
Now vary the policy without touching the paper. Under whole-letter bands, 86 is in the 80–89 range and is simply a B — same letter here, but a 91 would be an A under whole letters and an A− under plus/minus, so the two schemes agree less often than they appear to.
Suppose instead the test had 30 points and the student earned 26. Then 26 ÷ 30 = 86.667%. Left unrounded that is still a B. Rounded to the nearest whole percent it becomes 87, which is a B+ — one third of a grade point of GPA, decided entirely by the rounding rule. This is the single most common source of end-of-term grade appeals, and it is preventable by writing the rule down.
Finally, count errors instead of points. On a 20-question quiz with 3 wrong: 20 − 3 = 17 correct, 17 ÷ 20 = 85%, each question worth 5%. Notice that on this quiz no score of 86 or 87 exists at all — the possible grades are 85, 90, 95, 100 and so on downward. Short assessments cannot express fine distinctions, which is worth remembering before arguing about a single mark.
Reading the result as a teacher or as a student
As a student, treat a single test percentage as one input to a course grade, not as the course grade. What matters is the percentage multiplied by the assessment's syllabus weight. A 72% on a quiz worth 5% of the course costs you 1.4 percentage points of final grade; the same 72% on a final worth 30% costs 8.4. Run the numbers in the weighted grade calculator before deciding how alarmed to be, and use the final exam grade calculator to see what remains recoverable.
As a teacher, read the value of one question alongside the score. If each item is worth 5 percentage points, your grade resolution is 5 points and any scale with 3-point letter bands is finer than your instrument. Either lengthen the assessment, combine several into one weighted category, or accept that individual quiz letters are noisy and grade them on a coarse scale.
Watch the shape of the distribution rather than individual scores when you are judging whether a test worked. A single 43/50 says nothing about the test; twenty-five scores clustered between 40 and 46 say the test discriminated poorly at the top. The class average and distribution calculator gives you the mean, median, spread and letter counts, and the curved grade calculator shows what each standard curve would do about it.
One caution about percentages below 60. Because the F band spans 0 to 59 while every passing band spans about 10 points, percentage grading punishes a missed assignment far more than a poor one. A zero and a 50 are both failures, but the zero costs ten times as much in the average. That asymmetry is a known weakness of percentage grading and is the argument behind minimum-grade policies; whether to adopt one is a policy decision, not an arithmetic one.
Percentage and letter for a 20-question quiz
| Wrong | Correct | Percent | Plus/minus letter | Whole letter |
|---|---|---|---|---|
| 0 | 20 | 100% | A+ | A |
| 1 | 19 | 95% | A | A |
| 2 | 18 | 90% | A− | A |
| 3 | 17 | 85% | B | B |
| 4 | 16 | 80% | B− | B |
| 5 | 15 | 75% | C | C |
| 6 | 14 | 70% | C− | C |
| 7 | 13 | 65% | D | D |
| 8 | 12 | 60% | D− | D |
| 9 | 11 | 55% | F | F |
| 10 | 10 | 50% | F | F |
Notice that a 20-item quiz can never produce a B+ or a C+ under plus/minus bands: those bands are narrower than the 5-point gap between adjacent scores.
Mistakes that produce the wrong grade
- Counting questions when points differ. The questions-wrong shortcut is valid only when every item carries identical weight. A test with 2-point and 5-point items must be graded from points earned.
- Leaving bonus items in the denominator. Extra credit belongs in the numerator only; adding it to the points possible converts a bonus into a requirement.
- Rounding twice. Rounding each section of a test and then rounding the total can shift a paper across a band boundary that it never actually crossed.
- Assuming the 90/80/70/60 scale. Seven-point scales, where an A begins at 93 and a B at 85, are common in some regions and produce a different letter for the same percentage.
- Treating one test as the course grade. A test enters the course through its syllabus weight; on its own it predicts little.
- Grading a very short assessment on a fine scale. With 10 items your resolution is 10 percentage points, so plus/minus letters imply a precision the test cannot deliver.
Where a single test score goes next
A test percentage is an input to three other calculations. Multiplied by its syllabus weight it becomes part of the course grade, which the weighted grade calculator assembles across every category. Compared against a target it becomes a planning number, which the final exam grade calculator handles. And converted to a letter and then to grade points, it eventually reaches the GPA calculator and the cumulative GPA calculator by way of the course grade.
For teachers, the natural next step after grading a set is describing it. The class average and grade distribution calculator reports mean, median, standard deviation and the count in each letter band, which is what a department chair asks for. If the distribution is too low to stand, the curved grade calculator compares a flat addition, a square-root curve and scaling to the top score, and shows which students cross a boundary under each.
Two standards are worth naming even though neither dictates a cut-off. Criterion-referenced grading, which percentage scales implement, judges each student against a fixed standard; norm-referenced grading judges each against the class. Curving moves you from the first to the second, and doing so silently is the practice most likely to draw a complaint. State which model your course uses in the syllabus.
Finally, remember what a percentage cannot tell you. It has no information about which items were missed, whether the misses cluster on one objective, or whether the test measured what you meant it to. An item analysis answers those questions and a percentage never will — the number is a summary, and summaries discard exactly the detail that would let you teach differently next week.