Shafir E B, Smith E E, Osherson D N
Department of Psychology, Princeton University, NJ 08544.
Mem Cognit. 1990 May;18(3):229-39. doi: 10.3758/bf03213877.
The work of Tversky and Kahneman on intuitive probability judgment leads to the following prediction: The judged probability that an instance belongs to a category is an increasing function of the typicality of the instance in the category. To test this prediction, subjects in Experiment 1 read a description of a person (e.g., "Linda is 31, bright, ... outspoken") followed by a category. Some subjects rated how typical the person was of the category, while others rated the probability that the person belonged to that category. For categories like bank teller and feminist bank teller: (1) subjects rated the person as more typical of the conjunctive category (a conjunction effect); (2) subjects rated it more probable that the person belonged to the conjunctive category (a conjunction fallacy); and (3) the magnitudes of the conjunction effect and fallacy were highly correlated. Experiment 2 documents an inclusion fallacy, wherein subjects judge, for example, "All bank tellers are conservative" to be more probable than "All feminist bank tellers are conservative." In Experiment 3, results parallel to those of Experiment 1 were obtained with respect to the inclusion fallacy.
对于一个实例属于某一范畴的判断概率，是该实例在该范畴中典型性的递增函数。为了检验这一预测，实验1中的受试者阅读了对一个人的描述（例如，“琳达31岁，聪明，……直言不讳”），随后是一个范畴。一些受试者对这个人在该范畴中的典型程度进行评分，而另一些受试者则对这个人属于该范畴的概率进行评分。对于诸如银行出纳员和女权主义银行出纳员这样的范畴：（1）受试者认为这个人在联合范畴中更具典型性（一种合取效应）；（2）受试者认为这个人属于联合范畴的可能性更大（一种合取谬误）；（3）合取效应和谬误的程度高度相关。实验2记录了一种包含谬误，即受试者判断，例如，“所有银行出纳员都很保守”比“所有女权主义银行出纳员都很保守”更有可能。在实验3中，就包含谬误而言，得到了与实验1平行的结果。
