Test-retest reliability for multiple-target visual search: Eye-tracking and performance metrics

Birch-Hurst, Kayley; Cooper, Jamie; Foot, Isabella; Muckleston, Oli; Clark, Kait

Abstract

Visual search plays a critical role in life-saving careers such as radiology and airport security. There has been growing interest in exploring factors specific to individual observers (e.g., personality, strategy) that can affect search accuracy; however, paradigms and measures with low test-retest reliability are inappropriate for the exploration of individual differences. There is value in investigating individual differences in multiple-target search due to its relevance for professional searches, but the test-retest reliability of multiple-target search and its associated eye movements remains unknown. We tested a sample of undergraduate participants (n = 75) in two sessions (separated by 1-3 weeks) on a multiple-target search task (0, 1, or 2 target ‘T’ shapes, presented at high or low salience, amongst distractor ‘L’ shapes). We measured accuracy, response time, and eye movements and assessed performance across single-, dual-, and no-target trials. The results revealed a range of test-retest reliabilities: Intraclass correlation coefficients (ICC) for accuracy measures ranged from fair (.25 ≤ ICC < .4) to moderate (.4 ≤ ICC < .6), and more difficult-to-spot “low-salience” targets produced higher reliabilities than more easy-to-spot “high-salience” targets. Response time measures were overall more reliable than accuracy measures, ranging from moderate (.4 ≤ ICC < .6) to good (.6 ≤ ICC < .8). Eye-tracking measures (e.g., number of fixations, fixation duration, initiation time, saccade rate, saccade amplitude) had the best levels of reliability overall, ranging from good (.6 ≤ ICC < .8) to excellent (ICC ≥ .8), except for saccade duration which was moderate. These results indicate that only some multiple-target search metrics offer a strong basis for assessing individual differences and that future work should consider a multi-faceted approach to investigating visual search behaviour, integrating performance and eye-tracking measures.

• Presentation Conference Type: Poster
• Conference Name: European Conference for Visual Perception
• Start Date: Aug 24, 2024
• End Date: Aug 30, 2024
• Acceptance Date: May 31, 2024
• Deposit Date: Sep 10, 2024
• Print ISSN: 0301-0066
• Electronic ISSN: 1468-4233
• Publisher: SAGE Publications
• Peer Reviewed: Peer Reviewed
• Public URL: https://uwe-repository.worktribe.com/output/12881012