In Search of the Specificity and the Identifiability of Stochastic Thinking and reasoning

Abstract

Research on stochastic thinking and reasoning, with its implications on the teaching of statistics, should have professional input both from educational psychology as well as from the field of probability and statistics. Studying research papers on stochastic thinking and reasoning, one wonders which thought processes can be identified as being specific for the stochastic aspect of the experiments that are carried out. Such identification could be helpful, possibly leading to guidelines for optimizing statistical teaching strategies. Through a couple of examples, this paper tries to understand the complexity of apparently simple experiments. The chosen examples are very classical (about the “equiprobability bias” and about the “representativeness heuristic”), and many research papers have already appeared on these topics. It is shown that the mathematical complexity as well as the wording can easily play a confounding role when setting up experiments. An indication is given of how this could lead to new research questions.