Effects of aging and distractors on detection of redundant visual targets and capacity: do older adults integrate visual targets differently than younger adults?

Abstract

In the redundant target effect, participants respond faster with two (redundant) targets. We compared the magnitude of this effect in younger and older adults, with and without distractors, in a simple visual-detection task. We employed additional measures that allow non-parametric assessment of performance (Townsend's capacity coefficient) and parametric estimates (Linear Ballistic Accumulator model). Older participants' latencies were slower, especially in the presence of distractors, and their calculated capacity indicators increased with distractors. Parametric estimates indicated that these increases were generated by the older adults' increased difficulty in inhibiting the distractors, and not the results of either improved detection of redundant-targets, or of a generalized slowing of processing.

Figure 1. Response times for older and younger adults, across the two tasks (distractor-present and distractor-absent) and two types of target trials (single- and redundant-target).

Figure 2. Capacity coefficient for younger and older participants in distractor-present and distractor-absent tasks.

The thin dotted lines give C(t) estimates of each individual member of the pertinent group, and the thick solid line gives the group values, aggregated across all group members. C(t)  =  1 is the threshold for unlimited capacity – processing of one channel is unaffected by the presence of another target; C(t)> 1, super-capacity, implies that redundant-targets lead to superior performance; C(t) <1, limited capacity, implies that target processing in one channel is impaired by adding a target in the other channel.

Figure 3. The distribution of Houpt and Townsend's capacity test statistics, Cz, for younger and older adults across the two tasks (distractor-present and distractor-absent).

Distributions are plotted using violin plots. These plots consist of a box plot in the center of each violin, with a white circle representing the median, a black rectangle outlining the central 50% of the distribution, and a solid line extending to two standard deviations from the median. The grey area of the violin is a smoothed plot of the distribution of Cz values using a kernel density estimator. The dotted line represents the cut-off for statistically significant limited capacity.

Figure 4. Cumulative Distribution Functions for correct responses from redundant-, single- and no-target trials, averaged over younger and older participants in distractor-present and distractor-absent tasks.

Observed data are plotted using diamonds and model predictions using crosses.

Figure 5. Accumulation rate parameters for the parametric model of the redundant target paradigm.

The figure shows that in redundant-target trials, there are no effects for age and task-type. Yet in single-target trials, we observe a reduction in accumulation rates for older adults. Also note that accumulation rates for single-target trials are higher than for redundant-target trials, implying limited capacity.

Figure 6. A comparison of LBA parametric capacity estimates v_RT/v_ST, Panel A and redundant target effect (RTE) in ms, RT(redundant-target) – RT(fastest single-target), Panel B.

Comparing the two panels it is clear that the LBA parametric capacity follows the trend observed in the raw latency analysis. Mainly, adding a distractor to the task increases both measures, yet this increase is doubled for older adults.