Reactions of human subjects in simple sequential situations
The general purpose of this study was to explore the reactions of Ss to situations requiring a series of similar decisions. This was done within the framework of a mathematical analysis of situations; the framework owed much to game theory formulations. Particular purposes of the study were to observe the behaviour of individual Ss in a probability learning experiment, and in simple 2x2 games against nature.
The observations made were considered in the light
of some current theoretical notions about human behaviour in such situations. In particular, the stimulus sampling theory of Estes and his colleagues, the view of man as a processor of information according to Bayes' theroem, and the more general computer simulation views of behaviour were all examined. In general, neither stimulus sampling nor Bayesian accounts fit with the observations. All of the Ss studied were University students.
They react in a fairly lawful way. The reaction depends on the structure of the situation. Given some information, many Ss approach an appropriate reaction and some achieve it. Even with no information, some Ss approach an appropriate reaction. This seems to occur by the elimination of likely hypotheses about the situation and, finally, by the use of an elaborated set of rules paying attention to consecutive rewards or non-rewards. Observations were also made of the Ss' declared purposes and of their ability to recognise a sequence of binary events as a random one. Suggestions for further research were made.
|Sep 1, 1970