ABSTRACT
This chapter helps us to revisiting the basic binary decision task and introducing the linear model and the probability matching can be overcome by feedback and incentives. It also deals with extending the linear model to multiple-cue probability learning. The basic features of repeated decision learning with feedback are captured by the example of playing slot machines and receiving occasional payoffs. Probability matching, commonly observed in such situations as well as in one-shot decisions, is suboptimal but can be overcome when individuals are given enough feedback and adequate incentives. The standard probability-learning experiments reviewed in the last section can be thought of as special cases of this sort of task in which the number of symptoms is zero. The chapter concluded by remaining the attempts in a non-technical way, to show how it can be expanded to deal with situations involving multiple predictive cues.
