Population and sample
A population is the whole group you want to know about. A sample is the part you actually measure. A number describing the population is a parameter. A number from the sample is a statistic.
Last updated
Analogy
Tasting one spoonful of soup to judge the whole pot. The pot is the population, the spoonful is the sample.
Example
All voters in a country: population1,200 voters polled: sampleTrue share backing a law: parameterShare in the poll, 54%: statistic
In practice
Polls ask about 1,000 voters to learn about millions. Random selection makes the sample fair.
Practice
Flashcards
- Parameter or statistic: the mean of a sample?A statistic. A parameter describes the whole population.
Questions people ask
What is sampling bias?
Sampling bias happens when the way you pick people makes some groups more likely to be included than others. Surveying only gym members about exercise habits overstates how active people are. A simple random sample gives everyone an equal chance and avoids this problem.
Why does a larger sample not fix a biased sample?
Size reduces random noise but not systematic error. If you only ask gym members, a million answers still describe gym members. Bias pushes every estimate the same direction. A small random sample often beats a huge biased one, so fix how you pick people before you collect more.