Graduation Year


Document Type




Degree Name

Master of Arts (M.A.)

Degree Granting Department


Major Professor

Thomas Sanocki, Ph.D.

Committee Member

Chad Dube, Ph.D.

Committee Member

Geoff Potts, Ph.D.


Attention, Perceptual Learning, Repeated Exposure, Visual Perception


There is a set of visual processing advantages for holistic or global information over detailed or local information; these advantages are known as global precedence (Navon, 1977). Currently, there are inconsistent results about whether selective attention can reduce global precedence. Our studies look into Lamb et al. (1998)’s claim about selective attention’s inability to reduce global precedence. We reassess Lamb and colleagues’ claim by examining whether consecutively repeated tasks strengthen selective attention and reduce interference or facilitation from irrelevant information. Our studies utilized a series of trials, or runs, to present multiple consecutively repeated tasks. Before each run, participants were directed to a level (global or local) and target (A, E, G, K, U) and tasked to confirm or deny the presence of a target at a focused level. Inside each trial, participants were briefly shown one hierarchical letter - a large letter (one of the target letters) made up of small letters (one of the same five) to represent global and local levels respectively. The focused level always contained the target while the irrelevant level switched between congruent (e.g., giant A made of small A’s) and incongruent (e.g., giant E made of small A’s) information at specific points in a run (trials two or six). Response times (RT) to complete the task were analyzed. Our primary concern was the influence of the irrelevant level (i.e., interference or facilitation) during globally or locally focused runs (i.e., elevation in RT for an incongruent stimulus relative to congruent stimulus). Experiment 1 showed that local interference (i.e., influence of an incongruent local level during global searches) decreased from multiple consecutively repeated tasks but global interference was still large during local searches. Experiment 2 showed no local facilitation and an insignificant reduction in global facilitation. A fit to the power-function speed-up model used in Logan (1988) confirmed that participants switched from a general algorithm to instance-based strategies in our paradigm.

Included in

Psychology Commons