Effects of long -term memory accessibility on saccadic eye movements.

摘要

Saccadic eye movements vary in their rate depending upon the type of non-visual cognitive task. Ehrlichman et al. (2006) described "the long-term memory hypothesis" in which EMR directly reflects requirements for searching long-term memory in any non-visual cognitive task. The present study built upon this hypothesis to test whether accessibility of information to be retrieved from long-term memory (LTM) impacts eye movement rate (EMR). The primary goal was to assess whether low accessibility conditions would produce higher EMR than high accessibility conditions within a long-term memory task. Additional goals were to extend previous findings of high EMR for LTM tasks to newly learned episodic memory and include a working memory task that could be varied on the basis of difficulty.;Participants performed episodic (levels of processing) and semantic (verbal fluency) LTM tasks with two levels of accessibility, and, as a control for difficulty, a working memory (n-back) task with two levels of difficulty. Thirty-two undergraduate students participated (10 males, 22 females). The results gave equivocal support for the accessibility hypothesis. Newly learned episodic memory was associated with significantly higher rates of eye movement than working memory, and there was no difference in EMR between varying levels of difficulty for the working memory task. There was no difference in EMR between the varying levels of accessibility for the episodic memory task; however, a significant difference in EMR was found between the accessibility levels of the semantic memory tasks. In addition, the fluency tasks were compared by 15-second quartiles. There was a clear negative correlation between performance and EMR over time, with low EMR associated with high performance in the first quartile and high EMR associated with low performance in the last quartile for the semantic fluency task. In addition, three clusters of tasks were found which significantly differed in their EMR: a working memory task with a vigilance component, rote memory and working memory tasks without a vigilance component, and long-term memory retrieval tasks. The paper also reviewed possible areas of overlap between long-term memory, working memory, and eye movements in the brain.