Statistics: Statistical Literacy

Understanding Statistics: Statistical Literacy

Numbers become meaningful only when their context is clear. A claim that a test score rose by ten per cent sounds important, but the starting score matters. Moving from ten to eleven is a ten per cent rise, even though it is only one point.

This is the difference between relative change and absolute change. News reports often use relative change because it creates a larger sounding number.

Check the actual counts too. If a risk rises from one person in ten thousand to two people in ten thousand, the relative risk doubled, while the individual risk remains very small.

Averages can hide important differences inside a group. The mean uses every value, so a few unusually high or low values can pull it away from what most people experience. Income is a common example.

A small number of very high incomes can raise the mean income far above the income of a typical worker. The median is often more useful for describing a typical value in uneven data. Spread matters as much as the centre.

Two classes can have the same average score, while one class has scores clustered closely together and the other has scores spread widely. Looking for the range, or other measures of spread, helps reveal this difference.

Graphs can clarify data, but design choices can change the impression they give. A bar chart with its vertical scale starting close to the values can make a small difference look huge. A line graph can suggest a steady trend even when it shows only a few measurements.

Check the labels, units, time period, and scale before accepting the visual message. Watch for graphs that compare unequal groups or use totals when rates would be fairer. For example, a city with more people may have more accidents in total, yet a lower accident rate per person than a smaller city.

Studies use samples because measuring every person is usually impossible. A sample can give a useful estimate when it represents the wider population fairly. Random selection reduces some bias, but it does not remove uncertainty.

A margin of error describes the normal random variation expected from sampling. It does not fix a survey that missed important groups, used leading wording, or had many people refuse to respond. Correlation is another place where caution is needed.

When two patterns move together, a third factor may influence both. Ice cream sales and sunburns can rise during the same months because hot sunny weather affects each one.

When reading a claim, separate the data from the conclusion. Notice what the evidence directly shows, what it cannot show, and what extra information would be needed before making a decision.