01 · Learn the idea
Differences and totals
Subtracting compares two results directly. If one seedling grew 12 centimetres and another grew 7, the difference of 5 centimetres is the comparison.
Adding gives totals, such as the total rainfall over a week, which lets you compare one period with another even when the daily numbers are uneven.
02 · Build understanding
Averages
An average summarises repeated trials into one number. Add the values and divide by how many there are: 20, 22 and 24 have a total of 66 and an average of 22.
An average is more trustworthy than a single trial, because a high reading and a low reading pull against each other instead of standing alone.
(20 + 22 + 24) ÷ 3 = 22
03 · Build understanding
What an average hides
Two sets can share an average and be very different. Values of 21, 22 and 23 average 22, and so do values of 2, 22 and 42.
So report the spread as well, or at least the highest and lowest values. An average alone can make unreliable results look tidy.
04 · Build understanding
Checking a calculation
An average must lie between the smallest and largest values. If it does not, the arithmetic is wrong, and that check costs a moment.
Keep the unit through the calculation. Averaging centimetres gives centimetres, and an answer with no unit cannot be used.
Apply and explain
Your turn
Explain what subtracting and adding each let you compare. Calculate the average of 20, 22 and 24 and explain why an average is more trustworthy than one trial. Show with an example what an average can hide. Then give a check that catches an arithmetic error.
Use paper or explain your thinking aloud before opening the solution.
Check your reasoning
Subtracting compares two results directly, so if one seedling grew 12 centimetres and another grew 7, the difference of 5 centimetres is the comparison. Adding gives totals, such as total rainfall across a week, which lets one period be compared with another even when the daily values are uneven. The average of 20, 22 and 24 is found by adding them to get 66 and dividing by 3, giving 22. An average is more trustworthy than a single trial because a high reading and a low reading pull against each other rather than standing alone, so one odd measurement has less influence. An average can hide a great deal: 21, 22 and 23 average 22, and so do 2, 22 and 42, even though the second set is far less consistent. The spread, or at least the highest and lowest values, should therefore be reported alongside the average. A check that catches an arithmetic error is that the average must lie between the smallest and largest values. If it does not, the calculation is wrong. Keeping the unit through the calculation is also necessary, since averaging centimetres gives centimetres and an answer with no unit cannot be used.
- Subtraction and addition each matched to a comparison.
- Average of 22 computed with the reliability argument.
- A worked example of hidden spread and the range check.
Ready to practice?
Try the questions, get feedback, and review your reasoning.
Data and explanations practice →