inferential statistics
Inferential Statistics
The branch of statistics concerned with drawing conclusions about populations from sample data — hypothesis tests, confidence intervals, regression, and any method that quantifies uncertainty in an estimate.
When you'd see it: A/B tests, scientific studies, survey analysis, any situation where you're using a sample to say something about a larger population. 'We sampled 500 customers and found a 95% confidence interval of [4.2%, 5.8%] for the true conversion rate' is inferential.
Why it matters: Almost all important business decisions are based on incomplete data — a sample, an experiment, a subset of customers. Inferential statistics is the set of tools for quantifying how much you should trust conclusions drawn from that incomplete data.
Common mistakes: Treating sample results as if they were population facts. A conversion rate of 4.3% in a 200-person test is not 'the conversion rate' — it's an estimate with substantial uncertainty. The inferential layer makes that uncertainty explicit and should be reported alongside any sample-based claim.
Study this in BizTech Primer →