Article #21
Newsletter list 


Previous Article 

Next Article 


Why Are Some Judges Better Than Others?

Elise Weaver & Tom Stewart
Albany, NY

We presented a poster at the Society for Judgment and Decision Making meeting in New Orleans describing pilot work (37 subjects) for our NSF-sponsored project on individual differences in judgmental skill.

We hypothesized that predictors of judgment accuracy in a multiple cue probability judgment task would include skill in cue probability learning as measured by Chasseigne et al.'s MCPL task. In addition we tested whether the following would also predict accurate judgment: crystallized intelligence, fluid intelligence, and coherent judgment (as measured by the Linda task, avoidance of violation of probability rules in subjective probability estimates, and the Wason selection task).

While our initial model was not a good fit, the following paths were supported: Fluid intelligence predicted judgment accuracy, and MCPL skill also predicted accurate judgment, over and above the contribution of fluid intelligence.

In contrast, scores on tests of coherence did not predict judgment accuracy. In addition, we found it necessary to separate different kinds of coherence tests because these tasks were associated with different kinds of intelligence (Wason with crystallized intelligence, and probability rules with fluid). Finally, we found that a frequency format of a Bayes task was correlated with accuracy, though not tested in our structural equation model.

We are somewhat encouraged in our hypothesis that judgmental accuracy is not wholly a function of intelligence, but is also related to skill at MCPL. We were surprised, however, that coherence tests were unrelated to judgment accuracy, since we asked people to make probability judgments. These are only preliminary results, and we need to refine our measures and increase the sample size.

Contact Elise Weaver & Tom Stewart

Previous Article                Next Article

Newsletter list