Pre-test and post-test learning evaluation
Measure change without confusing it with causal impact
Pre-test and post-test designs measure the same outcome before and after an educational experience. A single group estimates observed change; a credible comparison group helps assess changes unrelated to the intervention. Raw change, standardised change and normalised gain are distinct statistics. Measurement alignment, attrition, test familiarity and regression to the mean shape their interpretation.
Use it when learning change matters and suitable measures can be administered at meaningful times, preferably with a comparison design that can address maturation, history and initial differences.
Strengths
- Makes starting points and individual change visible
- Can improve precision when baseline measures predict later outcomes
Limitations
- Single-group gains confound intervention with other changes
- Repeated tests can introduce familiarity and ceiling effects
Know the boundary
Normalised gain is not Cohen’s d, and a statistically significant within-group gain does not establish intervention effectiveness.