Children and Adults Both Learn Motor Sequences Quickly, But Do So Differently

Abstract

Both children and adults can learn motor sequences quickly in one learning session, yet little is known about potential age-related processes that underlie this fast sequence acquisition. Here, we examined the progressive performance changes in a one-session modified serial reaction time task in 6- and 10-year-old children and adults. We found that rapid sequence learning, as reflected by reaction time (RT), was comparable between groups. The learning was expressed through two behavioral processes: online progressive changes in RT while the task was performed in a continuous manner and offline changes in RT that emerged following a short rest. These offline and online RT changes were age-related; learning in 6-year-olds was primarily reflected through the offline process. In contrast, learning in adults was reflected through the online process; and both online and offline processes occurred concurrently in 10-year-olds. Our results suggest that early rapid sequence learning has a developmental profile. Although the unifying mechanism underlying these two age-related processes is unclear, we discuss possible explanations that need to be systematically elucidated in future studies.

Keywords: age-related; declarative sequence knowledge; fast sequence learning; fatigue; implicit sequence learning; online process; offline process; task pacing.

FIGURE 1

Experimental setup and protocol. Ten 6-year-old children, 13 10-year-old children (12 were included for data analyses), and 10 adults performed a foot stepping serial reaction time (SRT) task. Participants responded to each visual stimulus as quickly and accurately as possible by stepping to the spatially matched target on the floor. The stepping performance was measured by reaction time and movement time. In blocks 1–4, the visual stimuli followed sequence A (10 repetitions of 1423564215). In block 5, the stimulus followed sequence B (10 repetitions of 3615425214). Each number was associated with one spatially located square, but the numbers were not displayed to participants. Both sequences consist of 100 trials and each stimulus appeared 1500 ms after the preceding stimulus. After each learning block, participants had a 3-min rest. Participants were not told that the stimuli followed a sequence until they completed all six blocks.

FIGURE 2

(A) The mean reaction time for each block (BMRT). The gray area represents the block in which the stimuli follows a novel sequence. The BMRT depended on learning blocks and age. Adults and 10-year-olds were faster than the 6-year-olds, while there was a trend that 10-year-olds had the same BMRT as adults. All three groups learned sequence A, as revealed by decreased BMRT from blocks 1 to 4 and increased BMRT from blocks 4 to 5. Such learning did not result from motor improvements as the BMRTs in blocks 1 and 5 were the same; (B) changes in RT within and between blocks. The solid line represents the trend as RT progressively changes, estimated by a local weighted regression. Shaded areas are the standard errors of the trend. Circles represent RMRTs (i.e., the mean RT of one repetition of stimulus sequence). The adults’ RTs progressively decreased within each block, while there were offline boosts in RT in 6-year-olds; and (C) average online and offline RT changes when learning sequence A from blocks 1 to 4. Note that negative values imply decreases in RT (i.e., RT becomes faster). Both offline and online changes in RT relied on age. Specifically, the offline change was greater in 6-year-olds (greater than 0) than 10-year-olds and adults. The online change was greater in 10-year-olds and adults than 6-year-olds whose RT deteriorated ‘online.’ Error bars represent standard errors of the mean performance within each block. RT, reaction time.

FIGURE 3

(A) Learning measured in both the first and last 50 trials in blocks. Learning magnitudes were comparable between two halves across all three groups; (B) corrected offline learning after removing the RT decrement within the preceding block from the total offline gain depended on age. Importantly, the corrected offline RT change was significantly different from 0 in 6-year-olds. Error bars represent standard errors. RT, reaction time.

FIGURE 4

(A) The mean movement time (MT) for each block. The gray area represents the block in which the stimuli follows a novel sequence. All three groups had comparable MT. In addition, MT did not change from blocks 4 to 5 in all three groups, suggesting that MT does not represent sequence learning. (B) Changes from blocks 4 to 5, in terms of MT, measured in both the first and second halves. The changes were comparable between two halves across all three groups; changes in RT within and between blocks. (C) Average online and offline MT changes when learning sequence A from blocks 1 to 4. Note that negative values imply decreases in MT (i.e., MT becomes faster). Online changes in MT did not rely on age. In contrast, offline MT changes significantly depend on age. Specifically, the offline change was greater in 6-year-olds than adults; and (D) when the offline gain was corrected by removing the MT decrement within the preceding block from the total offline change, the age effect vanished. Importantly, the corrected offline MT change was no longer significantly different from 0 in 6-year-olds. Error bars represent standard errors of the mean performance within each block. MT, movement time.

FIGURE 5

(A) The recall score. There was a tendency that the recall score reduced with age decreasing; (B) recognition score. Six-year-olds showed lower recognition score compared to adults and 10-year-olds. Error bars represent standard errors.