Work-sample tests of trainability: A meta-analysis

Abstract

The difference between normal work-sample tests and trainability tests is described, in that the former are for trained people and the latter are tests that incorporate a learning period and are specifically for untrained people. Several meta-analyses were conducted using different predictor criterion pairs. Total N s = 2,542 and 2,772. The meta-analyses were based on independent samples on which information was available on sample size, validity coefficient, type of predictor and, in most cases, type of criterion and length of follow-up period. The results of the meta-analyses suggest that trainability work-sample tests predict short-term training success more accurately than longer term training success, but there may be greater situational variability when short-term follow-up periods are involved. (PsycInfo Database Record (c) 2025 APA, all rights reserved)