-experimenter evaluation/no criterion). In fact, the opportunity for self-evaluation alone led to performance equivalent to that produced when participants subject to experimenter evaluation were asked to do their best to generate as many uses as possible (experimenter evaluation/ no criterion). However, this output may not represent the participants' best performance. In nearly 400 studies involving 40,000 participants in eight countries, 88 different tasks, and time spans ranging from 1 minute to 3 years, Locke and Latham (1990a) reported that participants urged to strive to attain a specific, difficult level of performance did even better than did participants asked to do their best. They argue that this goal-setting effect is the result of the potential for self-evaluation. So, just as Szymanski and Harkins (1987) argued that for self-evaluation to be possible, participants must have access to some measure of output and a criterion, Locke and Latham (1990b) argued that the goal-setting effect requires that participants have knowledge of their output (or feedback in Locke and Latham's terminology). ...
