Data Types, Populations & Study Design

Statistics turns messy data into defensible decisions under uncertainty. You describe samples, quantify chance, and infer what is plausible about a larger population.

What statistics does

Statistics turns messy data into defensible decisions under uncertainty. You describe samples, quantify chance, and infer what is plausible about a larger population.

Population vs sample: the whole vs the subset you actually measure.
Parameter (population truth) vs statistic (sample number).
Categorical (labels) vs quantitative (numbers you can average).
Observational study vs experiment (random assignment enables causal claims).

Correlation is not causation — without a randomised experiment, confounders can fake a relationship.

Bias and sampling

Simple random sample: every subset of size n equally likely.
Stratified: sample within important subgroups.
Convenience samples bias toward whoever is easy to reach.
Non-response and undercoverage quietly warp results.