Attentional control constrains visual short-term memory: insights from developmental and individual differences

Abstract

The mechanisms by which attentional control biases mnemonic representations have attracted much interest but remain poorly understood. As attention and memory develop gradually over childhood and variably across individuals, assessing how participants of different ages and ability attend to mnemonic contents can elucidate their interplay. In Experiment 1, 7-year-olds, 10-year-olds, and adults were asked to report whether a probe item had been part of a previously presented four-item array. The initial array could either be uncued, be preceded ("precued"), or followed ("retrocued") by a spatial cue orienting attention to one of the potential item locations. Performance across groups was significantly improved by both cue types, and individual differences in children's retrospective attentional control predicted their visual short-term and working memory span, whereas their basic ability to remember in the absence of cues did not. Experiment 2 imposed a variable delay between the array and the subsequent orienting cue. Cueing benefits were greater in adults than in 10-year-olds, but they persisted even when cues followed the array by nearly 3 seconds, suggesting that orienting operated on durable short-term representations for both age groups. The findings indicate that there are substantial developmental and individual differences in the ability to control attention to memory and that in turn these differences constrain visual short-term memory capacity.

Figure 1. Experiment 1 – Trial schematic.

On “neutral” trials (central row of images), participants were simply presented with a four item array and were later asked to report whether a probe item had been part of the array. The probe was present in the current example. On “pre-cue” trials, the array was preceded by spatial cues (represented by the cartoon character at fixation trying to ‘help’ the participants by pointing to the location of the item in the array that he thought would subsequently be the probe item) orienting attention to one of the four array locations, where the probe item, if present, would be (in “valid” trials) or at one of the other three locations (on “invalid” trials). On “retro-cue” trials, cues followed the array. The final probe item remained visible until a response was made. Feedback for responses was provided in the form of engaging auditory inputs: if participants responded correctly they heard one of the following, depending upon the speed of response: >1000 ms “Good”; 800-1000 ms “Great”; 600-800 ms “Super-spectacular”; and, <600 ms “You’re on fire”. Incorrect responses resulted in the voice of the cartoon character saying “Oops”

Figure 2. Experiment 1 – Mean D-prime benefits and costs following valid and invalid pre-cues and retro-cues for 7-year-olds, 10-year-olds and adults.

Benefits and costs were calculated as the difference in d-prime scores between cued trials and neutral trials.

Figure 3. Experiment 2 – Trial schematic.

As in Experiment 1, participants reported whether the probe item had been present in the initial array. On “neutral” trials, the probe was presented either 2500 (“short neutral trials”) or 4500 ms (“long neutral trials”) after the onset of the array. On “retro-cue” trials, the array was followed by a spatial cue presented either 700 ms (“short retro-cue trials”) or 2700 ms after array offset (“long retro-cue trials”). Retro-cues always indicated the location of the probe accurately if this was present in the array.

Figure 4. Experiment 2 – Mean D-prime benefits following retro-cues by 10-year-olds and adults when cues followed the array by a short or long delay.

Benefits were calculated as the difference in d-prime scores between cued trials and neutral trials.