Did you know that two school groups can report average scores of 68 and 78, yet the combined average need not be 73? It depends on how many learners each average describes. This catches people because ‘average the averages’ sounds fair, but it gives a small group the same influence as a large one.
The combined mean is the sum of all individual values divided by the total number of individual values. When you know each group’s mean and size, rebuild each group’s total as group mean × group size, add those totals, and divide by the combined number of learners. This gives each individual observation one equal share. It is different from calculating the average of the two group means.
Quick answer: the three quantities you need
- For each group, write n (number of learners or observations), mean (average), and unit (marks, minutes, books or another consistent measure).
- Find each group total using n × mean; this reverses the arithmetic-mean calculation.
- Add totals and counts separately; never add averages without considering how many observations each represents.
- Divide combined total ÷ combined count and check the answer falls between the original means when group sizes are positive.
- If a group mean was rounded, report an approximate combined mean rather than pretending to know every original score.
Worked classroom example: a smaller group and a larger group
Entirely illustrative numbers, not actual eduKate pupils or assessment outcomes. Twelve learners in Group A have an exact mean score of 68 points on a 100-point quiz. Eighteen learners in Group B have an exact mean score of 78 points on that same quiz. Assume all 30 learners took the same assessment scored on the same scale, that no scores are missing and that the two stated means are exact.
Group A’s total points: 12 × 68 = 816 points. Group B’s total: 18 × 78 = 1,404 points. Combined total: 816 + 1,404 = 2,220 points. Combined learners: 12 + 18 = 30 learners. Therefore combined mean = 2,220 ÷ 30 = 74 points per learner.
Check by reversing: 74 × 30 = 2,220, exactly the reconstructed point total. The result is between 68 and 78 and nearer 78 because the higher-mean group contains more learners. The unweighted shortcut (68 + 78) ÷ 2 = 73 describes the mean of the two group means, not the mean score of all 30 learners.
Why does the ordinary shortcut give the wrong answer?
Think of every learner as carrying one score into the final total. If Group A has 12 learners, the value 68 represents twelve scores altogether. The value 78 represents eighteen scores. Averaging 68 and 78 directly behaves as if each group had the same size. It silently changes the question from ‘What is the average learner score?’ to ‘What is the average of the two groups as two units?’ Both quantities can be defined, but they answer different questions.
The principle is already contained in the definition of an arithmetic mean: total of data values divided by count. OpenStax Introductory Statistics, Measures of the Center, explains calculating means from a total and a frequency. Here group size is the frequency represented by a reported group mean.
A compact diagram you can draw without a formula
Make two boxes on paper. Box A says ’12 learners, 816 total points’. Box B says ’18 learners, 1,404 total points’. Put a plus sign between the totals and another between the counts. Finish with 2,220 points shared across 30 learners. Ask the student to explain where each number comes from. If they can only recite a formula, remove the formula and use the two-box representation again.
Diagnose the actual misunderstanding before assigning practice
- Shortcut error: The learner immediately calculates (68 + 78) ÷ 2. Ask which group has more learners and whose scores should appear more often.
- Wrong denominator: They divide 2,220 by 2. Ask what two counts represent and why the average is per learner rather than per group.
- Mixed units: They add a percentage mean to a raw mark mean without checking whether assessments use the same scale. Re-establish the variable and its unit.
- Rounded input error: They call a reconstructed total exact when reported averages were rounded. Ask whether the original total can be known exactly.
- Interpretation error: They think 74 means every pupil scored 74. Explain that the mean summarises a group and does not describe each individual’s mark.
A second worked case: why equal-size groups are special
Imagine two additional fictional groups, each with eight learners, mean scores of 60 and 80 points. Their totals are 8 × 60 = 480 and 8 × 80 = 640. Combined count = 16; combined total = 1,120; mean = 1,120 ÷ 16 = 70 points. The shortcut (60 + 80) ÷ 2 also gives 70, because both group sizes are identical. Equal group sizes make the two approaches agree. That is a condition, not a universal rule.
Another special case: if two groups have the same mean, combining them preserves that mean regardless of how different their sizes are. If five learners average 72 and fifty learners also average 72, the combined mean is 72. This is a useful independent check rather than a reason to abandon the total-over-count method.
Progressive practice: from guided to independent
Guided attempt: A club has four learners averaging six books and six learners averaging eleven books read. Write the units before the calculation. The totals are 4 × 6 = 24 books and 6 × 11 = 66 books. Combined: 90 books ÷ 10 learners = 9 books per learner. Ask the learner to explain why (6 + 11) ÷ 2 = 8.5 is not the combined club average.
Part-supported attempt: Two teams complete a skills drill: 8 participants average 45 seconds, while 12 participants average 55 seconds. The learner must supply the totals, denominator and unit. Correct check: (8 × 45 + 12 × 55) ÷ 20 = (360 + 660) ÷ 20 = 51 seconds per participant. The higher-mean team has more participants, so the combined mean must lie closer to 55.
Independent transfer: One reading group has 15 students averaging 40 minutes of independent reading and another has 5 averaging 60 minutes. Without hints, find the combined mean and explain whether it is nearer 40 or 60. Answer for checking: (15 × 40 + 5 × 60) ÷ 20 = 900 ÷ 20 = 45 minutes. Then invent a different class-size pair with the same means and predict how the combined mean moves.
What if the group averages were rounded?
Suppose a class report says ‘about 68 points’ and ‘about 78 points’ rather than providing exact means. Multiplying these approximate means by group sizes reconstructs approximate totals. A combined answer of about 74 is defensible for the original example, but 74.000 is not new precision. Ask for the original totals or unrounded group means when a precise overall result matters.
Similarly, do not combine means from unlike measures: ‘78%’ from one quiz and ’68 marks out of 80′ from another do not yet refer to the same numerical scale. Converting 68/80 to 85% puts both onto a percentage scale, but meaningful aggregation still depends on whether the comparison is of learners, assessments or weighted grading components. A mean across unlike tasks may not answer a useful question.
What parents and teachers can do in five minutes
Write three cards: a group mean, its size and its total. Have the learner arrange them into the story ‘people × mean = combined contribution’. Give a deliberately imbalanced pair (for example, 2 learners versus 18 learners) so the influence of group size becomes obvious. Praise a correct explanation of the denominator before speed. Next day, present the same structure as a reading-time or transport-time problem; transfer across context is the goal.
Track observable progress using a simple four-check record: identifies counts correctly; reconstructs totals correctly; divides by total count; interprets the answer with units and limits. A student may need support on only one of these steps. Repeating ten whole exercises without locating that step is rarely the most informative approach.
Frequently asked questions
Can I just average two averages? Yes when the groups are equal-sized, or when the question specifically asks for the unweighted mean of the two groups as group-level objects. Otherwise use observation counts.
Can I combine three or more groups? Yes. Multiply each exact mean by its count, add all the implied totals, and divide by the sum of all counts. No new principle is required.
Is this the same as a weighted exam grade? Not always. Here the weight is the number of equally counted observations. For an assessment with deliberately different course-component weights, use the published assessment weights instead.
Does the combined mean reveal individual variation? No. Two datasets can have the same mean but very different spreads. If variation matters, inspect the underlying data or suitable spread summaries.
Useful next reading
For a related denominator trap, see how different sample sizes change the meaning of research table columns. For the distinct question of time weighting, compare unevenly timed temperature readings. Browse the eduKate Education Hub or the Learning Hubs for the next subject-specific problem. The Secondary 1 Mathematics Clementi reference shows the diagnosis-first, explanation-and-checking approach used here.
The core aim is not to memorise an impressive formula. It is to protect the meaning of the quantity being averaged. If the learner can rebuild the total, name the denominator and independently test the result, the method becomes useful well beyond a school worksheet.