Skip to Main Content
 

Global Search Box

 
 
 
 

ETD Abstract Container

Abstract Header

Mechanisms of Visual Relational Reasoning

Hayes, Taylor Ray

Abstract Details

2014, Doctor of Philosophy, Ohio State University, Psychology.
Humans possess an extraordinary ability to extract relational information even in completely novel task environments. What are the underlying mechanisms that make this relational extraction process possible? The presented work investigates this question across 5 studies by studying how people solve visual analogy problems from a benchmark test of fluid intelligence, Raven's Advanced Progressive Matrices (RAPM). One study develops a novel method for extracting statistical regularities from complex sequences of eye movements (Successor Representation Scanpath Analysis, SRSA) to quantify the role of individual differences in attentional control during Raven performance. The results revealed that 41% of variance in RAPM score could be explained by individual differences in problem solving strategies. In a followup study SRSA was used to study RAPM practice effects commonly observed in the test-retest designs used in the cognitive enhancement literature. The findings revealed that RAPM practice effects can be explained by refinements in strategy, suggesting strategy refinement as a potentially serious confound in cognitive enhancement research that uses visual training regimens (e.g. visual N-back tasks). In two followup studies relational reasoning was studied by actively manipulating the order and amount of RAPM problem information. The findings established a causal relationship between strategy and rule insight and suggested participants deploy attention to one cell of the problem matrix at a time. Finally, a novel combination of pupillometry and verbal protocol analysis was used to understand how the mediation of the exploration-exploitation tradeoff contributes to individual differences in fluid intelligence. Converging evidence from primate electrophysiology and computational neural modeling have indicated that changes in exploratory versus exploitive control state may be mediated by the broad noradrenergic projections emanating from the locus coeruleus (LC). At the same time, pupil diameter has recently emerged as a promising non-invasive proxy measure for LC activity. The present study pioneers the use of pupillary response to understand how the exploration-exploitation tradeoff contributes to visual relational reasoning. The results showed an increase in pupil diameter during exploratory periods and decrease during exploitative periods consistent with prominent theories of LC function. In addition, 25% of the variance in RAPM score was explained by individual differences in pupillary response during exploratory periods. This is the first study showing evidence that noradreneric activity may mediate the exploration-exploitation tradeoff during relational reasoning and the only study to show an exploration-related boost in pupillary response that is modulated by individual differences in fluid intelligence.
Alexander Petrov (Advisor)
Roger Ratcliff (Committee Member)
Per Sederberg (Committee Member)
James Todd (Committee Member)
185 p.

Recommended Citations

Citations

  • Hayes, T. R. (2014). Mechanisms of Visual Relational Reasoning [Doctoral dissertation, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1416933187

    APA Style (7th edition)

  • Hayes, Taylor. Mechanisms of Visual Relational Reasoning. 2014. Ohio State University, Doctoral dissertation. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1416933187.

    MLA Style (8th edition)

  • Hayes, Taylor. "Mechanisms of Visual Relational Reasoning." Doctoral dissertation, Ohio State University, 2014. http://rave.ohiolink.edu/etdc/view?acc_num=osu1416933187

    Chicago Manual of Style (17th edition)