A college's reported rate rises from 20% to 25%. One description says it increased by five points; another says it increased by 25%. Both can be arithmetically correct, yet they describe different quantities. A reader who does not identify the unit can mistake a modest absolute change for a much larger shift in the student population.

OpenLearn's mathematics material distinguishes percentage-point differences from relative changes in percentages. The calculations below apply that distinction to an invented college indicator. They do not report an actual institution's improvement or identify the reason for any change.

Write the starting value next to the change

Assume the same defined outcome is recorded for two comparable fictional cohorts, each containing 400 students. In the first cohort, 80 meet the outcome; in the second, 100 do. Their rates are 20% and 25%. The absolute rate difference is 25 minus 20, or five percentage points.

Measure Calculation Result
First rate 80 ÷ 400 20%
Second rate 100 ÷ 400 25%
Percentage-point difference 25 − 20 5 points
Relative rate increase (25 − 20) ÷ 20 25%
Additional students meeting outcome 100 − 80 20

The relative increase uses the original 20% as its denominator. It does not mean that an additional quarter of all students met the outcome. That would be a twenty-five-point gain, taking a 20% rate to 45%. Including the two endpoint values makes the intended meaning much harder to misread.

Direction also matters. A rise from 20% to 25% is a 25% relative increase, but a fall from 25% back to 20% is a 20% relative decrease. The percentage-point distance remains five in either direction. Relative changes differ because the starting denominators differ.

Let the population size change

Now retain the rates but suppose the second cohort contains 200 rather than 400 students. Twenty-five percent of 200 is fifty. The rate rises while the number meeting the outcome falls from eighty to fifty. Neither statistic is wrong; they answer different questions about prevalence and scale.

This distinction is useful when comparing programme expansion, scholarship coverage or student participation. A larger count does not necessarily show a higher participation rate, and a higher rate does not necessarily mean more places or more people. Always retain the numerator and denominator where the publisher supplies them.

Check that the underlying outcome and observation period are the same. Changing from six-month employment to fifteen-month employment, or from all entrants to surveyed graduates, is not just a numerical update. The units may look identical while the populations or definitions differ. In that case the arithmetic difference can be computed but should not be described as an improvement on a consistent measure.

Repair a misleading headline

When a claim says only “up 25%,” find the original table and write the endpoints. If the publisher supplies only rounded percentages, avoid reconstructing exact student counts. A displayed 20% may represent a range of underlying values, so multiplying by a denominator can yield a count that was never reported.

For a decision worksheet, use four columns: earlier rate, later rate, percentage-point change and relative change. Add the cohort sizes separately. Mark the relative change undefined when the original rate is zero; a move from zero to a positive rate cannot be expressed as a finite percentage increase using this formula.

The strongest short description is explicit: “In this fictional equal-size example, the outcome rose from 20% to 25%, a five-percentage-point or 25% relative increase, representing twenty additional students.” The extra words are doing real work. They prevent three different quantities from being collapsed into a single impressive-looking number.