The effect of lifestyle on some non-communicable diseases: Students should be able to understand the principles of sampling as applied to scientific data in terms of risk factors.

The idea

Sampling means studying a subset of a population to estimate something about the whole population, such as how a lifestyle risk factor relates to a non-communicable disease. A sample must be large enough and selected without bias, for example choosing every tenth name from a list rather than only asking people in a gym. If the sample is small or biased, the estimate of risk may be unreliable. Scientists compare groups that differ in one risk factor, such as smokers and non-smokers, and look at disease rates. Correlation does not prove causation, because other factors may differ between groups. Understanding sampling lets you judge whether a claimed link between a lifestyle factor and a disease is trustworthy.

What earns the marks

• Defines a sample as a subset of a population selected to represent the whole population.
• Explains that sample size affects reliability: larger samples generally give more reliable estimates of risk.
• Explains that random or systematic selection reduces bias, whereas convenience sampling can distort results.
• Describes comparing groups that differ in one risk factor, such as smokers and non-smokers, to estimate associated risk.
• Recognises that correlation between a risk factor and a disease does not by itself prove causation.
• Identifies that other variables, such as age, diet or occupation, should be controlled or accounted for.

Common mistakes and how to put them right

• Assuming any sample represents the whole population.

  Put it right by checking size, selection method and whether the sample matches the population.

• Confusing correlation with causation.

  Put it right by stating that a third factor or chance may explain an observed link.

• Thinking a larger sample removes all bias.

  Put it right by explaining that a large but biased sample can still give misleading results.

Exam tips

• Name the sampling method and say why it reduces bias, rather than just writing 'it is fair'.
• Use the phrase 'the sample may not be representative' and then state a specific reason.
• When interpreting a risk comparison, quote the groups and the outcome, for example smokers and non-smokers with lung cancer rates.