Resumen

FERREIRA, Mário Boto  y  GARCIA-MARQUES, Leonel. O papel do reconhecimento do acaso no raciocínio indutivo. Aná. Psicológica [online]. 2003, vol.21, n.3, pp.353-373. ISSN 0870-8231.

Human inference is, according to Tversky and Kahneman (1974), greatly based in heuristics (simplified rules of decision making) that diverge from the statistical principles appropriate to judgment under uncertainty. However, Nisbett, Krantz, Jepson and Kun-da (1983), contend that beyond these non-statistical heuristics, people also possess statistical heuristics (i.e., intuitive counterparts of statistical principles). Fong and Nisbett (1991), suggest that the use of statistical heuristics depends from codification rules associated to specific content domains. In the present work, we propose that codification rules facilitate the recognition of the chance component inherent to inductive problems. Thus, experimental conditions that facilitate the recognition of this chance component should result in the improvement of statistical performance. To test this hypothesis, we used two manipulations meant to facilitate chance recognition. Participants answered to a number of inductive problems concerning diverse content domains (Sports, Fidelity, School and Health), and regarding a number of statistical principles (Law of the great numbers, Regression to the mean, Base-Rates, and Diagnosticy) before and after the manipulations (immediately and two weeks after). The results showed that statistical performance improved in the domains of Fidelity and School, and for problems concerning Regression to the mean and Diagnosticity principles.

Palabras clave : Judgment under uncertainty; heuristics.

        · resumen en Portugués     · texto en Portugués     · Portugués ( pdf )