Systematic variability in stage-one processing as revealed by eye movements in efficient visual search

Ng, Gavin Jun Peng

Permalink

https://hdl.handle.net/2142/101245 Copy

Description

Title

Systematic variability in stage-one processing as revealed by eye movements in efficient visual search

Author(s)

Ng, Gavin Jun Peng

Issue Date

2018-04-27

Director of Research (if dissertation) or Advisor (if thesis)

• Lleras, Alejandro
• Buetti, Simona

Department of Study

Psychology

Discipline

Psychology

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

M.S.

Degree Level

Thesis

Keyword(s)

• visual search
• efficient search
• eye movements

Abstract

Most models of visual search have reported that efficient search tasks are characterized by “flat” search slopes of less than 10ms/item. Based on this, such models have concluded that stage-one processing times are invariant. In contrast, our lab recently reported systematic variability in stage-one processing times. In an efficient search task with a fixed target, reaction times were observed to increase logarithmically with set size. In addition, this logarithmic increase was modulated by the visual similarity between the target and the lures. We thus proposed a new model of visual search: stage-one processing is characterized by a parallel, capacity-unlimited evidence accumulation process that is affected by both lure-target similarity (due to resolution limits of the eye) as well as the number of items on the display (due to the exhaustive nature of evidence accumulation). Here, we examined eye movements to further elaborate on these novel contributions of our model. When lure-target similarity was high, participants made more fixations, had longer initial saccade latencies, and were more likely to make an initial saccade that is target-uninformed. Effects of set size were also observed on some of these variables. Interestingly, these eye movement results are incompatible with our model’s proposal of a completely exhaustive stage-one process. To reconcile these differences, we update our model to propose that (1) top-down control factors such as strategy and task instructions influence the time to stop evidence accumulation in stage one and (2) exhaustive processing occurs at a smaller spatial scale (i.e. not necessarily across the entire display) although all locations are accumulating evidence.

Graduation Semester

2018-05

Type of Resource

text

Permalink

http://hdl.handle.net/2142/101245

Copyright and License Information

Copyright 2018 Gavin Ng

Owning Collections