Repeated measurements of spatial ability with the Manikin Test.

Abstract

A task reflecting spatial ability, the Manikin Test, was examined for parallelism of repeated measurements. (Only parallel test scores are suitable for repeated measurement of human performance.) The results bear on some theories of individual differences. Egan's theory that latency scores reflect spatial ability (whereas accuracy scores reflect something else) was supported, as was Jones's theory that individual differences stabilize with practice. The hypothesis of Baron and Treiman, that skill transfers between tests of the same ability, was not supported. The results illustrate the principle that not all tasks and scores are useful for repeated measures. The Manikin Test log latency score was shown to be suitable for repeated-measures applications.