Skip to main content

Research Repository

Advanced Search

Temporal Distinctiveness in Task Switching: Assessing the Mixture-Distribution Assumption

Grange

Temporal Distinctiveness in Task Switching: Assessing the Mixture-Distribution Assumption Thumbnail


Authors



Abstract

In task switching, increasing the response-cue interval has been shown to reduce the switch cost. This has been attributed to a time-based decay process influencing the activation of memory representations of tasks (task-sets). Recently, an alternative account based on interference rather than decay has been successfully applied to this data (Horoufchin et al., 2011a). In this account, variation of the RCI is thought to influence the temporal distinctiveness (TD) of episodic traces in memory, thus affecting their retrieval probability. This can affect performance as retrieval probability influences response time: If retrieval succeeds, responding is fast due to positive priming; if retrieval fails, responding is slow, due to having to perform the task via a slow algorithmic process. This account-and a recent formal model (Grange and Cross, 2015)-makes the strong prediction that all RTs are a mixture of one of two processes: a fast process when retrieval succeeds, and a slow process when retrieval fails. The present paper assesses the evidence for this mixture-distribution assumption in TD data. In a first section, statistical evidence for mixture-distributions is found using the fixed-point property test. In a second section, a mathematical process model with mixture-distributions at its core is fitted to the response time distribution data. Both approaches provide good evidence in support of the mixture-distribution assumption, and thus support temporal distinctiveness accounts of the data.

Acceptance Date Feb 8, 2016
Publication Date Feb 24, 2016
Publicly Available Date Mar 29, 2024
Journal Frontiers in Psychology
Publisher Frontiers Media
Pages 251 -?
DOI https://doi.org/10.3389/fpsyg.2016.00251
Keywords task switching, decay, interference, computational model
Publisher URL http://dx.doi.org/10.3389/fpsyg.2016.00251

Files




You might also like



Downloadable Citations