Did you know that the average of two accurate class pass percentages can be an inaccurate pass percentage for the combined classes? A small class and a large class cannot always be given equal weight when we are counting how many students passed.
The immediate answer: to combine pass percentages, add the number who passed in each non-overlapping group, add the number eligible in each group, then divide the two totals. Equivalently, use a weighted average of the percentages with class sizes as the weights. Taking the simple mean of the displayed percentages answers a different question: the average percentage of the two classes, treating each class equally.
First diagnose the denominator
Ask the learner this short question: Class A has eight students, six of whom passed. Class B has forty students, twenty of whom passed. Is the school-wide pass percentage (75% + 50%) ÷ 2 = 62.5%? An answer of 62.5% reveals that the learner has averaged the class labels without counting the people behind them.
Now ask: ‘How many students passed, altogether? How many students were included?’ There were 6 + 20 = 26 passes out of 8 + 40 = 48 students. The correct combined percentage is 26 ÷ 48 × 100%, approximately 54.17%. It is nearer the larger class’s 50% because that class contributes five times as many students.
Why the simple average gives the wrong whole-school story
The number 75% describes six of eight pupils; 50% describes twenty of forty. Percentages are ratios. They summarise how large one count is relative to its own denominator. A plain mean of 75 and 50 treats the two classes as if they were the same-sized units, although the first percentage describes eight students and the second describes forty.
When the question is ‘What proportion of all 48 students passed?’, each student counts once. Class A contributes eight student records and Class B forty. A class-level average can be an appropriate measure for a different purpose—for example, comparing the typical classroom’s pass rate—but the result must be named honestly. The denominator defines the statistic.
- Class A: 6 pass / 8 enrolled = 75%.
- Class B: 20 pass / 40 enrolled = 50%.
- Unweighted average of two class rates: (75% + 50%) / 2 = 62.5%.
- Overall student pass rate: (6 + 20) / (8 + 40) × 100% = 54.166…%, or 54.17% rounded to two decimal places.
- Weighted check: (8 × 75% + 40 × 50%) / 48 = 54.166…%. Both valid routes agree.
All figures above are invented teaching numbers, not actual school results. They assume the groups do not overlap, every listed student has the same pass/fail definition and all forty-eight students are counted once. Change those assumptions and the appropriate calculation may change.
The mechanism: each percentage has an invisible weight
The weighted-average formula is (n₁p₁ + n₂p₂) / (n₁ + n₂), where n is the relevant group size and p is the group’s proportion or percentage, used consistently. For these classes, n₁ = 8, p₁ = 0.75, n₂ = 40 and p₂ = 0.50. Therefore (8 × 0.75 + 40 × 0.50) ÷ 48 = (6 + 20) ÷ 48. Converting this proportion to a percentage gives approximately 54.17%.
A useful second route is simply to recover the numerator: class size multiplied by pass proportion equals the number who passed, provided the displayed percentages are exact rather than heavily rounded. If class percentages were rounded beforehand, the original counts are more reliable than reconstructing passes from those rounded figures.
For the underlying distinction among proportions, ratios and rates, see Statistics Canada’s data-literacy guide. For the general principle of calculating means from values and frequencies, see OpenStax Introductory Statistics.
A new worked example: the library reading challenge
Illustrative scenario. Two school reading clubs report that a book-review challenge was completed by 9 of 12 members in Club P and 14 of 28 in Club Q. Assume one participation record per pupil, identical completion criteria and no pupil in both clubs. What percentage of the forty members completed the challenge?
- Club P: 9 ÷ 12 = 0.75 = 75%.
- Club Q: 14 ÷ 28 = 0.50 = 50%.
- Combined completed: 9 + 14 = 23 members.
- Combined eligible: 12 + 28 = 40 members.
- Combined completion rate: 23 ÷ 40 × 100 = 57.5%.
- Independent check: (12 × 75 + 28 × 50) ÷ 40 = (900 + 1,400) ÷ 40 = 57.5%.
The plain class average would again be 62.5%, overestimating the combined completion rate by five percentage points. Notice that a percentage point difference is found by subtraction: 62.5% − 57.5% = 5 percentage points. A relative percentage difference would answer a different question.
Three tempting mistakes, and quick repairs
Mistake 1: adding percentages across different denominators
Writing 75% + 50% = 125% completed is not a meaningful combined pass rate here. The two groups do not share the same denominator. First restore the counts or use a correctly weighted formula.
Mistake 2: applying weights that represent the wrong thing
If Club P ran twice as many sessions as Club Q, using session counts as weights would not answer ‘What proportion of pupils completed?’ Weight by eligible pupils for that question. For a time-weighted temperature average, durations rather than student counts are appropriate. The unit counted depends on what the statistic is meant to describe.
Mistake 3: assuming a single overall percentage tells every story
The 57.5% overall number is correct for the combined group, but it hides the different 75% and 50% club rates. Keep both subgroup and overall figures when the comparison matters. Do not claim that class size caused the difference in completion; the numbers describe outcomes, not their causes.
Progressive practice that builds independence
Guided: equal class sizes
Two groups of 20 students have pass rates 40% and 70%. Their pass counts are 8 and 14. The combined rate is 22/40 = 55%, which also equals the simple mean (40% + 70%)/2. Ask why this shortcut happens to work: the two denominators are identical.
Supported: unequal class sizes
A class of 10 has 90% completion, and a class of 30 has 60%. Convert to 9 and 18 students, combine to 27/40 and obtain 67.5%. Compare with the misleading unweighted 75%. Explain why 67.5% is closer to 60%.
Independent: answer without the formula beside you
A club has 15 of 20 members complete a task; another has 24 of 30 complete it. A: find each club percentage. B: find the combined completion percentage. C: explain whether the simple mean of the club percentages happens to match. D: state one condition needed to interpret the totals honestly.
Answers: A = 75% and 80%; B = (15 + 24)/(20 + 30) = 39/50 = 78%; C = 77.5% does not match because group sizes differ; D might mention non-overlapping members, consistent definitions or no missing denominator records.
How to see progress in schoolwork and at home
A learner can put a small label beneath every percentage: numerator/denominator. A teacher can show two bar charts with the same class percentages but different class sizes and ask whether the whole-school story changes. A parent can ask, ‘Out of how many?’ whenever a report shows one percentage without context.
Two days later, use new counts and check whether the student independently reconstructs the total numerator and denominator instead of averaging displayed percentages. True progress appears when the method transfers to attendance, survey response, reading challenges and sports participation without a memorised rule being prompted.
Limits and useful cautions
Never merge percentages representing incompatible events or populations: a pass rate and an attendance rate do not have a common numerator; different eligibility rules may make a pooled figure misleading. If a source provides only rounded percentages and no denominators, an exact pooled rate may be impossible to recover. If the same person appears in two groups, adding group totals double-counts them. This is arithmetic literacy, not permission to infer why performance differs.
Frequently asked questions
When is it fine to average two percentages normally?
If each percentage has equal weight for the question being asked—for example, two equal-sized non-overlapping groups when calculating an overall proportion. It may also be legitimate to report an average class percentage when each class is the unit of interest; make that distinction explicit.
What if there are three or more groups?
Add every relevant numerator and every relevant denominator. The equivalent weighted average uses all group sizes. The principle does not change when additional groups are added.
Is this the same as the median of medians?
No. Combining medians requires different reasoning, and group medians alone generally do not identify the whole-group median. The shared lesson is to respect what information each summary preserves and what it loses.
What the learner should remember
Before combining any percentages, find out who or what each denominator counts. The arithmetic is simple once the question is clear. Combine counts for a people-weighted result, retain subgroup figures when they matter, and label the statistic so readers know exactly what it describes.
Further reading: why the median of class medians is not the school median, how a time-weighted average differs from a simple mean, the eduKate Education Hub, and our Clementi Mathematics explanation-and-practice reference.