descriptive
Descriptive statistics
Use when the outcome/model family is not yet clear or the needed model is not implemented.
Requirements
- clean variables
- clear research question
- appropriate model family
Avoid when
- a runnable inferential test is already appropriate
Formula
x̄ = (1/n) Σxᵢ; s² = Σ(xᵢ − x̄)² / (n − 1)
Interpretation
Summaries and plots guide cleaning/model choice; this is not an inferential test.
Worked example
Summarize age, glucose, and treatment group before choosing an inferential model.
Open example datasets