multiple tests

Terms from Statistics for HCI: Making Sense of Quantitative Data

A form of cherry picking where multiple tests are performed on different aspects of the same data but only the successful ones are reported. For example, a few tests that show some effect may be selected and the rest ignored, or tests that show a desired effect may be selected while those that indicate the opposite are ignored.

Defined on page 79

Used on pages 65, 79, 80, 82