Influences Aided and Unaided Visual Search Performance: Effects of Secondary Task Modality, Secondary Task Processing Code, and Personal Preferences

The present study examined the impact of secondary task modality and secondary task processing code on aided and unaided performance for a simulated luggage screening task. The interaction between secondary task modality and processing code with automation was examined. Finally, the interaction between processing preferences (as measured by learning styles) and outcome measures was assessed. One hundred sixtyone undergraduate students from Old-Dominion University performed a simulated luggage screening task over the course of two days. Half of the participants (n = 81) had the assistance of an automated aid called DETECTOR, and half did not. Participants also performed one of four possible secondary tasks in parallel with the primary screening task: visual-verbal, visual-spatial, auditory-verbal, auditory-spatial. All participants were assessed for sensitivity, criterion setting, average confidence, and workload ratings. Aided participants were further assessed for system trust, perceived reliability of the aid when the target was present and absent, compliance, reliance, confidence when agreeing with the aid, and confidence when disagreeing with the aid. Results revealed that aided participants had higher sensitivity, lower criterion settings, higher average confidence, and higher workload ratings than unaided participants. When comparing undistracted to distracted performance, visual distractors were detrimental to performance (as measured by sensitivity), and aided and unaided participants performed differently depending on the secondary task processing code. Results are discussed in relation to multiple resource theory, arousal theory, and automation design/development.