This review considers experimental research that has used probability theory and statistics as a framework within which to study human statistical inference. The experiments have investigated estimates of proportions, means, variances, and correlations, both of samples and of populations. In some experiments, parameters of populations were stationary; in others, the parameters changed over time. The experiments also investigated the determination of sample size and trial-by-trial predictions of events to be sampled from a population. In general, the results indicate that probability theory and statistics can be used as the basis for psychological models that integrate and account for human performance in a wide range of inferential tasks.

This experiment investigates the relation between subjective probability and mathematical probability. Bayes' theorem provides the correct revision of probabilities as a result of new data. New data, however, may come in different sized samples. This experiment measured the accuracy of subjective probability revision as a function of sample size. The results show that accuracy decreases as sample size increases; the function is negatively accelerated. The gain in amount of data processed by increasing the sample size is at the expense of accuracy.Behavioral decision theory provides a framework for evaluating the extent to which selected behavior of human 5s corresponds with ideally consistent behavior as outlined by statistical decision theory (Edwards, 1961). The two primary variables of decision theory are probability and value; the corresponding psychological variables are subjective probability and utility. The present experiment studies the correspondence between subjective probability and mathematical probability.Recent developments within statistical decision theory focus on the correct revision of probabilities in the light of new information-the problem of revising the probability of a hypothesis as a function of the occurrence of a relevant datum (e.g., see

2 experiments presented Ss with sequences of data that 1st favored 1 hypothesis, and then changed to favor a 2nd hypothesis. After each datum Ss became more or less sure of which hypothesis was correct. They reflected this change of opinion with probability estimates, which were compared with probabilities calculated by means of Bayes's theorem. Estimated probabilities changed from favoring the 1st hypothesis to favoring the 2nd hypothesis later than did corresponding Bayesian probabilities. Data that occurred early in a sequence influenced Ss more than did later data-a primacy effect. This result agrees with results of comparable experiments on impression formation.