Literature Index

It is not unusual today - even among people who consider probability as a concern of pure mathematics - to start a probability course with an attempt to uncover the experimental roots of the probability concept. In fact, it is not a new feature. The story about tossing a coin with the happy result of a fair distribution of heads and tails in the long run has been the custom for quite a long time. What is new about it, is that the story is dramatized and acted out - I mean, by the author or the textbook. Maybe even the teachers or the students are expected to try out this experiment - following the highly encouraging examples given by the textbook authors. It is a pity that by showing one experiment without asking themselves whether it is typical, textbook authors lead the teachers up the wrong path and help to create wrong attitudes towards probabilistic problems.

The NSF-funded Quantitative Literacy Project (QLP), a joint project of the American Statistical Association (ASA) and the National Council of Teachers of Mathematics (NCTM), served as the basis from which the strand in statistics was developed for the NCTM Standards. The QLP provides curriculum materials in certain areas of data exploration, probability, and inference, in a style that makes the material accessible to teachers and students, and provides a model framework for in-service programmes to enhance the skills of teachers in the area of statistics and probability. More specific information on certain aspects of the QLP will be provided below.

Every day the social perceiver makes numerous, apparently complex social judgments - Predicting another's behavior, attributing responsibility, categorizing an individual, evaluating anothers, estimating the power or influence of a person, or attributing causality. A central task of social psychology has been to determine how the social perceiver makes these judgments. Until recently, research on this topic was marked by a rationalistic bias, the assumption that judgments are made using thorough, optimal strategies (see, for example, Fischhoff, 1976, for discussion of this point). Errors in judgment were attributed to two sources: (a) accidental errors due to problems with information of which the perceiver was presumably unaware; and (b) errors which resulted from the irrational motives and needs of the perceiver. However, over a period of years, a growing body of evidence suggested not only that people's judgments and decisions are less complete and rational than was thought but that not all errors can be traced to motivational factors. Even in the absence of motives, judgments are often made on the basis of scant data, which are seemingly haphazardly combined and influenced by preconceptions (see, e.g., Dawes, 1976). These findings led to a revised view of the cognitive system. People came to be seen as capacity-limited, capable of dealing with only a small amount of data at a time. Rather than being viewed as a naive scientist who optimizes, the person was said to "satisfice" (Simon, 1957) and use shortcuts that would produce decisions and judgments efficiently and accurately.

The article attempts to sketch a conceptual and experimental history of the base rate issue. The review distinguishes between a social judgment paradigm and a textbook paradigm. Theoretical explanations of the base rate phenomenon, i.e. the representativeness heuristics, confusing or inverting conditional probabilities, the specific factor, the causality factor, the vividness factor etc. are discussed with respect to these paradigms. On the basis of a report on various studies, the author hypothesizes that the different paradigms elicit different response tendencies, matching base rates on the one hand and judging representativeness on the other hand. She argues that on the causal structures accepted illustrating this by several examples. The consequences of this phenomenon for the role of normativeness and its (in-) determinacy are discussed.