Date

2015

Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Psychology

First Adviser

Arrington, Catherine M.

Other advisers/committee members

Chein, Jason; Hupbach, Almut; O'Seaghdha, Padraig

Abstract

Working memory is one system that supports the flexible nature of cognition. It is a highly debated topic, with several competing models of organization. One recent model (Oberauer, 2009) is progressive, drawing on past work with well-supported adaptations. The model describes two subsystems, declarative (representing items that can be manipulated such as lists of words and numbers) and procedural (representing procedures used for manipulation, like processes for mental math) working memory. Both include three components with successively higher levels of activation. The base layer is an activated portion of long term memory, followed by a capacity-limited portion from which candidates for attentional selection are drawn (the bridge and region of direct access), and a focus of attention that highlights a particularly relevant item. One assumption of this model is that declarative and procedural working memory are functionally analogous, which means that they may operate according to equivalent mechanisms. However, much of the work investigating this model has explored declarative information. This dissertation used the task and memory span paradigms (Logan 2004) to evaluate this selected assumption of functional analogy in Oberauer's (2009) model.Experiment 1 established the task span as a test of procedural working memory, a novel methodological contribution. Experiment 2 tested the fate of recently activated information in the intermediary working memory stores. Experiment 3 tested the characteristics of information capacity in the intermediary procedural store. All three experiments shared methodology, the task span, in different variations. This dissertation tested the strong assumption that the two subsystems are functionally analogous. Results from Experiment 1 indicate that the task span is a suitable methodology for testing procedural working memory. Experiment 2 was inconclusive because of failure to replicate Oberauer's past work possibly due to floor effects in the memory span. Experiment 3 yielded results that are consistent with Oberauer's model. Results from all experiments are discussed in the context of Oberauer's theoretical model and a computational version of that model. In all, the current data provide support for the assumption of a functional analogy, and leave open questions of independence and of mechanisms of declarative and procedural subsystem communication.

Included in

Psychology Commons

Share

COinS