Universität zu Köln

Esser, Sarah (2017). The Role of Unexpected Events in the Emergence of Explicit Knowledge in an Implicit Learning Situation. PhD thesis, Universität zu Köln.

Preview

PDF DissertationSarahEsser.pdf Download (1MB)

Abstract

In an implicit learning task like the serial reaction time task, most people demonstrate implicit knowledge about the underlying regularity. Usually, a small group of persons can be found which notices this regularity and is also able to report it. Whether the acquired representation can be used in a flexible and diverse way crucially depends on conscious awareness of this knowledge. Therefore, it is of great importance to understand which mechanisms realize the transition from implicit to explicit knowledge. Research on this issue has led to two main theoretical streams. On the one hand, the most parsimonious account assumes that unconscious representations gain quality through practice and therefore gradually transform into explicit knowledge (single-system account; e.g. Cleeremans & Jiménez, 2002). On the other hand, there are more complex models which assume that implicit and explicit representations are supported by separable learning- and memory systems (multiple-systems account). One of these models is the Unexpected Event Hypothesis (Frensch et al., 2003). Within this model, it is proposed that implicit learning leads to behavioral changes which contradict the expectations of a person about their own behavior in the given situation. This violation of expectations triggers an attributional process which should bring expectation and experience back into coherence; a sudden insight into the underlying rule can be the result. The three studies presented here are aimed at testing the predictions of the Unexpected Event Hypothesis and contrast these with the more parsimonious predictions of a single system account. Therefore, in all three studies, different manipulations will induce unexpected events in an implicit learning situation. At the same time, all tasks are designed in a way to match the associative strength of the representations between the manipulations.