ABSTRACT
In the present dissertation I develop the PD-SOP model, a mathematically formalized
associative model of learning. The origins of the PD-SOP model are based on
Wagner's (1981) SOP model and, more directly, on the revised version of Wagner's
SOP, developed by Dickinson and Burke (1996). However, because those models
are path independent (i.e., they are unable to store the associative history
of the stimuli), they are also unable to account for the different effects of
response recovery that are due to several kinds of manipulations performed after
the learning stage (e.g., introduction of retention intervals or contextual
switches). As I will show in the present dissertation, PD-SOP model will be
able to account for a wide set of evidences coming from the associative learning
literature (in both non-human and human subjects). Concretely, the exclusive
features of PD-SOP as a path dependent model will enable it to explain those
effects of response recovery, which remained unexplained by its predecessor
models, as well as by most of the contemporary associative models of learning.
(in Spanish only, sorry)