Methods
Cognitive testing
An overview of how cognitive tests are designed, validated, and interpreted — and the pitfalls every reader should know.
Key takeaways
- Reliability, validity, and standardisation are the three foundations of any test.
- Norm samples matter as much as test items.
- Practice effects, motivation, and test-taking strategy all affect scores.
Reliability and validity
A reliable test gives consistent scores across administrations. A valid test measures what it claims to measure. They are independent properties; a test can be reliable but invalid.
Reading test results
Percentile rank is usually more interpretable than raw score. Confidence intervals matter. Single scores rarely tell you what people want them to tell you.