What is the difference between descriptive and inferential statistics?

Instruction: Explain both concepts and give an example of how each would be used.

Context: This question tests the candidate's knowledge of the two main branches of statistics and their ability to apply this knowledge.

Updated

Official answer

Descriptive statistics summarize the observed data: counts, means, medians, spread and plots. Inferential statistics use a sampling or probabilistic model to estimate unknown quantities or test hypotheses, with uncertainty.

Suppose 200 survey respondents include 120 who prefer a feature. Saying “60% of these respondents prefer it” is descriptive. Estimating the preference proportion among all eligible users, with an interval, is inferential and requires a defensible sampling design and handling of nonresponse.

A larger convenience sample can still be biased. Confidence intervals quantify uncertainty under a model; they do not automatically correct an unrepresentative sample. The same method, such as regression, may be used descriptively, predictively or inferentially depending on the goal and assumptions.

State which population and quantity the conclusion concerns. Describing the collected sample accurately is useful even when broader inference is unsupported.

Related Questions