Files in this item



application/pdfSteelman-Allen_Kelly.pdf (3MB)
(no description provided)PDF


Title:Modeling the control of visual attention in complex workspaces
Author(s):Steelman-Allen, Kelly S.
Director of Research:McCarley, Jason S.
Doctoral Committee Chair(s):McCarley, Jason S.
Doctoral Committee Member(s):Kirlik, Alex; Kramer, Arthur F.; Morrow, Daniel G.; Wickens, Christopher D.
Department / Program:Psychology
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Visual Attention
Models of Attention
Display Design
Abstract:Attentional behavior in complex visual workspaces is driven by the physical and temporal characteristics of the display, the goals and knowledge of the operator, and task demands. Accordingly, to develop effective displays, designers must understand how these factors interact to influence attentional allocation within the display and the detectability of critical alerts. The purpose of the current project was two-fold. First, an integrative model of attention was developed to serve as a theory-motivated design tool, allowing a designer to investigate the interactions between the many factors that influence attentional allocation and noticing behavior. Second, to fill in gaps in the existing theoretic literature and to support model development, the current project provided a pair of empirical studies examining the interactions between bottom-up and top-down sources of attentional guidance on alert detection and attentional behavior to inform the design of information displays. Built on the framework of SEEV (Wickens et al., 2003; Wickens & McCarley, 2008; Wickens et al., 2008), the model integrates elements from several basic attentional processes to create a model of attentional behavior in dynamic environments. The model was validated against PDTs from a high-fidelity flight simulator study, against miss rates and RTs from Nikolic et al.’s (2004) alert detection experiment, and data from the current studies. Combined, the results indicated that the model can accurately predict attentional behavior within complex environments and tasks, using heuristic parameter values selected by either the experimenter or an SME. The two current experiments examined how alert expectancy interacts with salience, eccentricity, and workload to influence attentional behavior in an alert detection task alert. Participants performed a central highway-in-the-sky (HITS) manual tracking task and a peripheral alert detection task concurrently while alert eccentricity, salience, expectancy, and central task workload were manipulated. The results of the studies suggested two primary design recommendations: 1) Critical, but low expectancy, alerts should be placed close to the central display, but away from other dynamic display items. 2) The location of high expectancy alerts should be determined by the importance of the alert. Alerts that require more prompt responses by the operator should be located closer to the central display and/or presented away from other dynamic display elements.
Issue Date:2011-08-25
Rights Information:Copyright 2011 Kelly S. Steelman-Allen
Date Available in IDEALS:2011-08-25
Date Deposited:2011-08

This item appears in the following Collection(s)

Item Statistics