An integrated measure of display clutter based on feature content, user knowledge and attention allocation factors. Issue 5 (4th May 2018)

• Record Type: Journal Article
• Authors: Pankok, Carl; Kaber, David B.
• Is Part Of: Ergonomics. Volume 61:Issue 5(2018)
• Page Start: 682
• Page End: 696
• Publication Date: 2018-05-04
• Subjects: Display clutter -- eye tracking -- driver behaviour -- visual search -- interface design; Human engineering -- Periodicals; Cybernetics -- Periodicals; Industrial management -- Periodicals; Ergonomie -- Périodiques; Cybernétique -- Périodiques; Gestion d'entreprise -- Périodiques; 620.8205
• DOI: 10.1080/00140139.2017.1401669
• Languages: English
• ISSNs: 0014-0139

Abstract

Abstract: Existing measures of display clutter in the literature generally exhibit weak correlations with task performance, which limits their utility in safety-critical domains. A literature review led to formulation of an integrated display data- and user knowledge-driven measure of display clutter. A driving simulation experiment was conducted in which participants were asked to search 'high' and 'low' clutter displays for navigation information. Data-driven measures and subjective perceptions of clutter were collected along with patterns of visual attention allocation and driving performance responses during time periods in which participants searched the navigation display for information. The new integrated measure was more strongly correlated with driving performance than other, previously developed measures of clutter, particularly in the case of low-clutter displays. Integrating display data and user knowledge factors with patterns of visual attention allocation shows promise for measuring display clutter and correlation with task performance, particularly for low-clutter displays. Practitioner Summary : A novel measure of display clutter was formulated, accounting for display data content, user knowledge states and patterns of visual attention allocation. The measure was evaluated in terms of correlations with driver performance in a safety-critical driving simulation study. The measure exhibited stronger correlations with task performance than previously defined measures.