作者: William H. Mertz , Michael E. Doherty
DOI: 10.1016/0030-5073(74)90046-4
关键词: Correlation 、 Clinical judgment 、 Linear model 、 Cognitive psychology 、 Test (assessment) 、 Social psychology 、 Additive model 、 Variance (accounting) 、 Task (project management) 、 Psychology
摘要: Abstract Many investigators have concluded that cue-combination in clinical judgment can be described by a simple, additive model. In this study cue characteristics were manipulated to see whether subjects might appear linear under some conditions but not others. Each of nine experienced high school guidance counselors made 1368 predictions college success based upon either one or two test scores. The two-cue data examined using several models. Although the model accounted for proportion predictable variance all conditions, within condition (high subjective correlation between cues), seven markedly nonlinear. It was cannot talk about configural judges independently task conditions.