Abstract
There are performance costs in search for two targets compared with two independent single-target searches. This dual-target cost remains after practice and with complex stimuli (Menneer et al., in press). Eye-movements during dual-target search for simple colour-shape conjunctions revealed a breakdown in colour selectivity that allowed the examination of irrelevant distractors (Cave et al., 2007). We examined the dual-target cost with complex stimuli and across training. Participants completed 16 blocks (5376 trials) within 40 days on search of baggage x-ray images. Twenty participants conducted single-target search, either for metal-threat items (guns and knives) or for improvised explosive devices (IEDs). Another group conducted dual-target search for both target types simultaneously. Eye-movements were recorded to determine the frequency of fixation to objects of each colour (orange, green, blue/black, mixed). Compared with single-target search, there were more fixations in dual-target search overall, as well as more time and more fixations taken to fixate the target. These differences held throughout training. In all searches, the majority of fixations were made to objects of the same colour as the predominant target colour(s): blue/black for metal-threats and mixed or orange for IEDs. In block 1, the proportion of fixations made to objects of non-target colour (green) was higher in dual-target search than in combined single-target search, but there was no difference in later blocks. In conclusion, more objects are fixated in dual-target search than in single-target search even after practice. There is an initial guidance breakdown in dual-target search as observed by Cave et al.. However, with training, the effectiveness of target representations appears to improve, reducing the examination of non-target-coloured objects, and equating to that in single-target search. The persistent dual-target cost therefore seems to arise from the number of fixations required rather than through guidance breakdown.
Supported by the Human Factors Program of the Department of Homeland Security's Transportation Security Laboratory.