Parameter vs Statistic: The Difference with Examples

A parameter is a number that describes a whole population. A statistic is a number that describes a sample. You usually can’t know the parameter, so you calculate the statistic and use it as an estimate.

A quick way to tell them apart

Ask: was this number calculated from every member of the group being described? If yes, it is a parameter. If it came from some of them, it is a statistic. The two words even share a first letter with what they describe: parameter and population, statistic and sample.

StatementParameter or statistic?
The mean age of all 1,200 employees at a firm, from HR records, is 38.4.Parameter. Every employee is included.
In a poll of 1,000 voters, 52% support the proposal.Statistic. It is from a sample of voters.
The 2020 US Census counted 331.4 million people.Parameter, at least in intent. A census tries to count everyone.
A batch test of 40 light bulbs found a mean life of 1,180 hours.Statistic. The parameter is the mean life of all bulbs made.

Statistics vary, parameters don’t

Draw a second sample of 40 bulbs and you will get a different mean, maybe 1,165 or 1,203 hours. The parameter stays put. This is the whole reason inferential statistics exists: statistics bounce around from sample to sample, and we need to know how much, so we can say how far our estimate might be from the truth.

The amount a sample mean typically bounces is the standard error, s ÷ √n. If the 40 bulbs had a standard deviation of 90 hours, the standard error would be 90 ÷ √40 = 14.2 hours. A rough 95% range for the true mean is the sample mean plus or minus about two standard errors: 1,180 ± 28.5, or 1,151.5 to 1,208.5 hours. The confidence interval calculator gives the exact interval.

Symbols

Parameters get Greek letters (μ, σ, ρ). Statistics get Latin letters (x̄, s, r) or a hat (p̂, read “p-hat”). The hat means “estimate of”. When you see a hypothesis like H₀: μ = 100, it is always about the parameter. You never test a hypothesis about a statistic, because you already know its value.

Check yourself

A hospital reports that the mean waiting time of all patients seen last month was 47 minutes. Parameter or statistic?

A parameter, if the population is “patients seen last month”. It becomes a statistic the moment you use it to talk about waiting times in general, such as next month’s.

Why is H₀: x̄ = 50 a meaningless hypothesis?

x̄ is calculated from your data, so you already know whether it equals 50. Hypotheses are about the population: H₀: μ = 50.