ACARA v9 CONTENT DESCRIPTION “investigate techniques for data collection including census, sampling, experiment and observation, and explain the practicalities and implications of obtaining data through these techniques”
Population and sample
The population is the entire group you want to know about. It might be every student in a school, every battery coming off a production line, or every household in a town. Often you cannot reach the whole population, because it is too large, too spread out, or too costly to measure in full. A sample is a part of that population, chosen and studied so that what you learn from the part can stand in for the whole. When you measure a sample you do not get the exact truth about the population, but a careful sample gives a sound estimate. The key idea is simple: the population is everyone of interest, and the sample is the smaller group you actually examine.
A population and a sample
Every dot is one member; a ringed block is a sample.
the population is the whole group of interest; a sample is a part of it chosen to learn about the whole.
Four ways to collect data
There are four common ways to collect data, and naming them clearly helps you choose well. A census measures every member of the population, leaving no one out. Sampling measures only a selected part of the population and uses it to estimate the whole. Observation records what naturally happens, watching and noting without stepping in to change anything. An experiment goes further: it deliberately changes one condition and then measures the result, so you can see what that change does. The first two are about how much of the group you measure; the last two are about whether you simply watch or actively intervene.
Four ways to collect data
Census, sampling, observation and experiment.
data can be gathered by a census, by sampling, by observation, or by experiment.
Census or sample?
A census gives complete and exact information, because nothing is left out, but it can be expensive, slow, or even impossible to carry out. Testing every battery until it fails would destroy the whole stock and leave none to sell, so a full census there makes no sense. A sample is faster and cheaper, since you measure far fewer items, but it only estimates the population and carries sampling variation: two different samples can give slightly different answers. Choosing between them means weighing completeness against cost and time. When the population is small and easy to reach, a census can be the better choice; when it is huge or measuring is costly, a sample is usually the sensible one.
Census or sample?
Complete and costly against fast and cheap.
a census is complete but costly and slow; a sample is faster and cheaper but only estimates the whole.
Observation and experiment
Observation and experiment differ in one clear way: whether the researcher steps in. Observation watches without interfering. Counting how many cars pass an intersection in an hour, or noting which products shoppers pick up, gathers data while leaving the situation untouched. An experiment introduces a change and then compares groups. To test whether a fertiliser raises crop yield, you might treat one group of plots and leave another group, the control, exactly as before, then measure the difference between them. The control group is what makes the comparison fair, because it shows what would have happened without the change. The deciding question is always the same: did the researcher change something, or only record what was already going on?
Observation and experiment
Watch without changing, or change and compare.
observation records what happens without intervening; an experiment changes something and measures the effect.
A representative sample
A sample is only useful if it mirrors the population in the ways that matter. If you want to know what a whole school thinks but ask only the members of one sports team, your sample is biased: it over-represents one group and leaves others out, so the conclusion can be badly wrong. A representative sample reflects the mix of the population, so its results can be trusted to stand for the whole. Drawing the sample at random, so that every member has an equal chance of being chosen, is one good way to help a sample stay representative; the particular methods for doing this are the focus of the next unit. For now the point is that a biased sample, drawn from one unrepresentative corner, quietly misleads.
A representative sample
A matching mix against a one-corner sample.
a representative sample mirrors the population; a biased sample over-represents one group and misleads.
Why this matters
Almost all real data comes from one of these techniques. Opinion polls, factory quality checks, medical trials, and market research each rely on a census, a sample, observation, or an experiment, and the choice shapes how far the results can be trusted. Knowing the difference between a population and a sample, and between watching and intervening, lets you read statistics critically and design fair data collection of your own. The most common slip is treating a biased sample as if it spoke for everyone, and once you can spot that, a great deal of misleading data falls apart. In this unit we stay with the four techniques and the population and sample idea; the methods of sampling, the effect of sample size on variation, and full statistical investigations come in the units that follow.
Quick self-check
1. What is the difference between a population and a sample?
2. Measuring every single member of the group of interest is called...
3. Why might a sample be used instead of a census?
4. What makes a study an experiment rather than an observation?