López-Romero LJ, Alvarado A, Tamayo C, Vila J

Language: Spanish
References: 0
Page: 113-120
PDF size: 220.57 Kb.

ABSTRACT

Recent studies had showed that animals choose between patches base on time using a dynamic average of experiences (Mazur, 1996). This average has been used to predict performance based on the temporal distance and the relative value of learning experiences. To test these ideas in humans we used a task in which participants could choose between two options with similar reinforcement densities over 5 trials (0.5 vs. 0.5) but different in the last or first trials block (0.9 vs. 0.1) And also they differed in training-test interval (0h vs. 24h). When reinforcement density in last training block was higher, a test at 0h showed preference for the option with higher density. Whereas when higher density ocurred during first trial block results showed indifference for both response options, but in a test at 24 hours they showed preference for the first higher option. These results are in agreement with Temporal Weighting Rule predictions and provide evidence for a dynamic average of past experiences in human choice.