Overview
Hypothesis testing fails when the test is chosen before the assumptions are checked. A t-test applied to non-normal data with small samples produces p-values that are wrong. A chi-squared test applied to cells with expected counts below 5 produces invalid results. A test that achieves p < 0.05 with N=50,000 may be detecting an effect so small it has no practical consequence.
The right hypothesis test is determined by the data structure (continuous/categorical/ordinal), the comparison type (two groups/multiple groups/paired/independent), the sample size, and whether the distributional assumptions are met. Each combination points to a different test — and a different interpretation of the result.
The Hypothesis Testing Framework Prompt generates a complete testing specification: assumption checks, test selection by data structure and comparison type, effect size calculation, power analysis, and a results interpretation that separates statistical from practical significance.
What you get: - Assumption diagnostic protocol - Test selection matrix by data structure and comparison type - Effect size calculation and interpretation - Power analysis for sample size determination - Results interpretation distinguishing statistical from practical significance
Built for: analysts, data scientists, and researchers who need defensible, assumption-checked hypothesis tests with interpretations that drive decisions.