Choosing the Right Average
- Your job here is to pick the average that best represents a dataset and to say why.
- Before you choose, look at the data for extreme values or for categories.
- What you spot decides which measure you reach for.
- An extreme value points you to the median, categories point you to the mode, and clean numerical data lets you trust the mean.
A Screen-Time Study with One Extreme Value
- A psychologist records yesterday's screen time, in hours, for eight teenagers: 2, 3, 3, 4, 4, 4, 5 and 22.
- Most of them are on about 4 hours, but one gamer is on 22.
- The median, the middle value once the scores are in order, is 4 hours, close to what most of them actually did.
- The mean is higher, pulled up by that single 22-hour reading, so it overstates the typical screen time.
- Report it like this: "The median of 4 hours is the best average here, because the one reading of 22 hours pushes the mean up and makes it misleading."
- That one sentence names the measure, gives the figure and gives the reason, which is exactly what the question wants.
When Categories Force the Mode
- Now suppose the same psychologist asks each teenager to name their main coping strategy: talking, exercise, music or gaming.
- These are categories, not numbers, so you cannot add them or find a middle; only the mode works, and if "music" is chosen most, that is the answer.
- Do not reach for the mean out of habit; check for an unusually high or low value, such as the 22, first.
- Do not state a measure with no reason; a choice without a justification earns no marks in an applied question.
A Step-by-Step Routine
- Scan the data for an extreme value or for categories.
- Pick the measure that gives the fairest typical value, and justify your choice in one sentence.
- For data with one extreme value, which average is usually best, and why?
- Which average best represents 1, 2, 2, 3 and 40, and why?
- A study records favourite coping strategy. Which average must you use?
- Why is a chosen measure worth little without a reason?