Response-level processing during visual feature search: Effects of frontoparietal activation and adult age

David J Madden,Zachary A Monge,Rachel E Siciliano,Catherine W Tallman,Andreas Voss,Jessica R Cohen

doi:10.3758/s13414-019-01823-3

Abstract

Previous research suggests that feature search performance is relatively resistant to age-related decline. However, little is known regarding the neural mechanisms underlying the age-related constancy of feature search. In this experiment, we used a diffusion decision model of reaction time (RT), and event-related functional magnetic resonance imaging (fMRI) to investigate age-related differences in response-level processing during visual feature search. Participants were 80 healthy, right-handed, community-dwelling individuals, 19–79 years of age. Analyses of search performance indicated that targets accompanied by response-incompatible distractors were associated with a significant increase in the nondecision-time (t0) model parameter, possibly reflecting the additional time required for response execution. Nondecision time increased significantly with increasing age, but no age-related effects were evident in drift rate, cautiousness (boundary separation, a), or in the specific effects of response compatibility. Nondecision time was also associated with a pattern of activation and deactivation in frontoparietal regions. The relation of age to nondecision time was indirect, mediated by this pattern of frontoparietal activation and deactivation. Response-compatible and -incompatible trials were associated with specific patterns of activation in the medial and superior parietal cortex, and frontal eye field, but these activation effects did not mediate the relation between age and search performance. These findings suggest that, in the context of a highly efficient feature search task, the age-related influence of frontoparietal activation is operative at a relatively general level, which is common to the task conditions, rather than at the response level specifically.

Highlights

Electronic supplementary material The online version of this article contains supplementary material, which is available to authorized users.In visual feature search, the target differs from all of the distractors in some dimension, whereas in conjunction search the target is a combination of distractor features (Treisman, 1988, 2004)
The analysis of variance (ANOVA) of drift rate yielded a significant effect of task condition, F(2, 79) = 8.51, p < 0.0003, ηp2 = 0.097
We hypothesized that the age-related differences in feature search, the effect of response-incompatible distractors, would be associated with nondecision time rather than drift rate, and that task-related activation would mediate the relation between age and search performance

Summary

Introduction

Feature search performance is highly efficient, as defined by minimal differences in reaction time (RT) as function of the number of items in the search display (Wolfe, 2014). This increased efficiency may reflect various mechanisms, including an early-stage parallel search of the display, as originally proposed by Treisman (1980), and the degree of similarity between the target and distractors (Duncan & Humphreys, 1989; Treisman, 1991) and the influence of top-down attention (Müller, Heller, & Ziegler, 1995; Wolfe, Butcher, Lee, & Hyle, 2003). Performance of conjunction search tasks becomes less efficient, as reflected in increased RT-display size slopes and

Methods

Results

Discussion

Conclusion