GENERALIZING FROM ATYPICAL CASES - HOW GENERAL A TENDENCY

We performed two studies to investigate the conditions under which ind ividuals use anecdotal (case history) versus statistical (base rate) i nformation in making judgments. In Study 1, a fictitious (unfamiliar) automobile, the Clipper, was rated after a 2-part essay about it was r ead that was either an anecdote followed by contradictory statistics o r statistics followed by a contradictory anecdote. Statistical informa tion that rendered the anecdote atypical had a great impact, whereas a necdotal information had little. In Study 2, 3 automobiles that varied in prior knowledge and evaluation were rated: Clipper (unfamiliar/neu tral), Honda (familiar/positive), and Yugo (familiar/negative). The im pact of statistical information on participants' evaluations was great regardless of prior evaluation or familiarity. We propose that the ex perimental paradigm used provides a good context for studying basic an d applied issues relating to people's use of base rate information, an d the results indicate that base rate information will be used when it is comprehensible and diagnostically relevant.