How To Calculate A Test Score

A test score begins as a raw count: the marks earned out of the marks available. Answer 42 questions correctly on a 50-question paper and the raw score is 42 out of 50. Divide the first number by the second and you get 0.84, which is the percentage score once it is multiplied by 100 to read 84 per cent. Those two figures describe the same performance at different scales, and confusing them is the most common error in the whole topic. The third figure, the grade, is not arithmetic at all.

The percentage is a fraction, not a grade. A mark of 0.84 is 84 per cent and a mark of 0.86 is 86 per cent; on the same 50-question paper those come from 42 and 43 correct answers respectively. Neither number tells you whether the result is a pass, a B or an A. That judgement belongs to a grade band, a separate institutional table mapping ranges of percentages onto labels, and it is precisely where the variation between schools, states and countries actually lives.

Raw scores also depend on how much each question is worth. When every question carries one point, 42 correct answers produce 42 points out of a maximum of 50. When questions are worth two points, the same 42 correct answers produce 84 points out of 100, and the percentage is unchanged at 0.84. This is why the points-per-question value belongs inside the calculation rather than in your head: it changes the raw and maximum totals while leaving their ratio alone, so a correctly handled paper gives the same percentage however the points are distributed.

The grade band is the part you cannot derive from the paper. A score of 0.84 might be recorded as a B at one institution and a B+ or an A- at another, because each awarding body publishes its own cut points. A difference of a single question is often enough to cross a boundary: 42 correct gives 0.84 while 43 correct gives 0.86, and if the A boundary sits at 0.85 then that one mark decides the letter. Read the band from the syllabus or the transcript rather than assuming a familiar scale applies.

The number of wrong answers is simply the total minus the correct count, which is why a 50-question paper with 42 correct has 8 wrong. Wrong answers matter for review but not for the percentage, because the percentage is fully determined by the two totals. It is still worth computing them as a check: 8 wrong plus 42 correct must add back to 50, and any mismatch signals an arithmetic slip rather than a marking dispute. Two of the three reported figures come straight from this subtraction.

Most school percentages terminate neatly, but many do not. Forty of fifty gives exactly 0.8, forty-five of fifty gives exactly 0.9, and forty-one of fifty gives exactly 0.82, so those convert without any loss at all. A ratio such as 40 out of 47 does not, producing 0.851063829787234 and forcing a decision about how many digits to keep. Carry the full value through any further calculation and round once, at the end, to the precision the grade band actually distinguishes. Truncating 0.8499 to 0.84 and rounding it to 0.85 is exactly the distinction that decides which side of a boundary a score of 42 out of 50 falls on.

Real papers are rarely uniform. A section worth two points per question and another worth one means the raw total is a weighted sum rather than a simple count, and the points-each value does the weighting. Multiply each section's correct count by its point value, add the products, and divide by the maximum built the same way. Ignoring the difference is how a paper with 30 easy one-point questions and 10 hard three-point questions ends up scored as though all 40 questions were equal, quietly understating the heavy section.

A percentage score and a percentile rank are not the same thing and should never be substituted for one another. A score of 0.84 says you answered 84 per cent of the paper correctly; a percentile rank says what share of the group you outperformed, which could be far higher or far lower depending on the cohort. Being 12th out of 250 is a strong percentile result, yet the percentage behind it might be anything at all. Keep the two calculations in separate columns so neither drifts into the other.

Test scores feed three practical decisions: whether a course is passed, what letter goes on the transcript, and how the mark combines with others in a course grade. Each of those steps applies its own rule on top of the raw division, so 42 out of 50 can contribute differently to a weighted course total than its 0.84 alone suggests. Compute the percentage cleanly first, then apply the course's own weighting and rounding conventions in the order the syllabus states. Reporting 0.84 where the band wanted 84 is a presentation error rather than a mathematical one, but it causes just as much confusion.