Studies examining ageing and skill acquisition in visual search have found conflicting results. Some researchers have found that when a semantic category search task is consistently mapped (CM), younger adults show significant disruption at reversal whereas older adults do not. This suggests that only younger adults are able to reach automatic levels of processing and thus, it has been hypothesised that older adults have difficulty manipulating the attention-attraction strength of targets and distracters. However, Scialfa, Jenkins, Hamaluk, & Skaloud, (Journal of Gerontology, under review) used a more traditional visual search procedure and found no age differences in disruption at reversal.

The present study examines this issue by incorporating eye movement data to investigate age differences in the acquisition of visual search skills in conjunction search. The stimuli in the display consisted of white and black line segments that were either oriented 45 deg to the right or to the left. The initial target was a white line oriented to the right and at reversal, the target was changed to a black line oriented to the right. Older (n = 10) and younger (n = 10) adults performed 3840 trials of CM conjunction search over 16 sessions. Reversals occurred every 4th session to determine the degree of disruption over time and to determine if transfer was evident over several reversals. Reaction time (RT) and eye movement data were recorded to measure search efficiency and to determine if participants could attain automatic levels of performance. Eye movements also determined whether participants could selectively attend to objects that possessed target-like features, suggesting some degree of rule-based transfer.

Consistent with Scialfa et al. (under review), results suggested that although age differences were apparent initially, these age differences were eliminated as training progressed. Reduction in both RT data and fixation frequency data demonstrated improvement in search efficiency as training progressed. Age differences in display-size effects continued throughout the training sessions. However, at reversal, RT and fixation frequency increased comparably for both age groups, suggesting no age difference in the degree of automatic processing. Furthermore, subsequent training sessions found no age differences in transfer. Results are discussed in terms of the priority-deficit learning hypothesis and generalised slowing.