| Abstract: | In 3 experiments with 72 undergraduates, Ss were asked to evaluate student performance based on 2 sources of information differing in variability. In Exp I the relative variability of test scores and paper grades assigned to a given student had no effect on the weighting of information, and evaluations were based on averaging the 2 types of information. In Exps II–III the weighting of test scores was compared for a highly variable grader and a less variable grader. Differential weighting occurred only when instructions linked variability to reliability. It is concluded that evaluators can assign less weight to a source of information perceived to be less reliable but they will not ordinarily use relative variability of scores as an index of reliability. (15 ref) (PsycINFO Database Record (c) 2010 APA, all rights reserved) |
