Humans as an animal model? Studies on cue interaction, occasion setting, and context dependency

The objective of the present thesis was to study human learning behaviour and to compare the results with those from animal learning studies. Three topics originating from animal learning research were examined: cue interaction, occasion setting, and context dependency. A series of experiments was first carried out to examine the influence of spatial position on cue-interaction effects in a predictive-learning task. Evidence that previously learned information about a stimulus can interact with learning about the predictive value of a second stimulus was only obtained under the condition that the stimuli were presented in a fixed spatial position. The next experiments examined the ability of a stimulus to modulate responding to other, mostly ambiguous, target stimuli (occasion setting). The temporal relationship between a potential occasion setter and a target stimulus was varied in the first experiment; the second experiment assessed whether there is a difference between children and adults in the associative process underlying modulation of responding to a target. The first experiment revealed that a stimulus is more likely to acquire occasion-setting properties when it precedes the target stimulus than when it is presented simultaneously with it. The second experiment revealed a larger propensity of adults than children to use the potential occasion setter as a true modulator of target responding rather than as a stimulus directly evoking a response by itself. The last series of experiments assessed the influence of a change in context from learning to testing on task performance. An easy predictive-learning task was used in the first two experiments; task difficulty was varied in the last two experiments. The results were that a context change resulted in impaired performance only if the task was sufficiently difficult. The collective results are indicative of a large overall correspondence between animals and humans in fundamental associative learning processes