Repetita iuvant: repetition facilitates online planning of sequential movements

Abstract

Beyond being essential for long-term motor-skill development, movement repetition has immediate benefits on performance, increasing speed and accuracy of a second execution. While repetition effects have been reported for single reaching movements, it has yet to be determined whether they also occur for movement sequences, and what aspects of sequence production are improved. We addressed these questions in two behavioral experiments using a discrete sequence production (DSP) task in which human volunteers had to perform short sequences of finger movements. In experiment 1, we presented participants with randomly varying sequences and manipulated 1) whether the same sequence was repeated on successive trials and 2) whether participants had to execute the sequence (Go) or not (No-Go). We establish that sequence repetition led to immediate improvements in speed without associated accuracy costs. The largest benefit was observed in the middle part of a sequence, suggesting that sequence repetition facilitated online planning. This claim was further supported by experiment 2, in which we kept a set of sequences fixed throughout the experiment, thus allowing participants to develop sequence-specific learning: once the need for online planning decreased, the benefit of repetition disappeared. Finally, we found that repetition-related improvements only occurred for the trials that had been preceded by sequence production, suggesting that action selection and sequence preplanning may not be sufficient to reap the benefits of repetition. Together, these results show that repetition can enhance representations at the level of movement sequences (rather than of individual movements) and facilitate online planning.NEW & NOTEWORTHY Even for overlearned motor skills such as reaching, movement repetition improves performance. How brain processes associated with motor planning or execution benefit from repetition, however, remains unclear. We report the novel finding of repetition effects for sequential movements. Our results show that repetition benefits are tied to improved online planning of upcoming sequence elements. We also highlight how actual movement experience appears to be more beneficial than mental rehearsal for observing short-term repetition effects.

Keywords: motor planning; repetition effects; sequence production; skill learning.

Fig. 1.

The discrete sequence production (DSP) task. A: experiment (Exp.) 1 example trial: a sequence cue (white numbers on the computer screen) is followed by a production cue (outline changes color, numbers are masked). Online visual feedback about keypresses was given during the movement phase (green asterisks for correct presses, red for incorrect presses), followed by reward points depending on performance. Thirty percent of the trials in a block were No-Go trials (red outline + low-pitch sound; top), 70% were Go trials (green outline + high-pitch sound; bottom). B: the next trial could be either a Repetition of the same sequence (0.5 probability) or a Switch to a new sequence. C: example trial in Exp. 1 with the following trial timing: preparation phase, 2.5 s; movement phase, 2 s; intertrial interval (ITI), 0.5 s. Dashed horizontal line indicates force threshold (1 N) to determine the moment of each keypress and release (dotted vertical lines). P1, press of first key; R4, release of fourth key; IPI1, first interpress interval. Total time (TT) = reaction time (RT) + movement time (MT). D: Exp. 2 design: 8 repeating sequences, trial structure, and timing. The go signal is given via a white box around the sequence cue.

Fig. 2.

Immediate repetition leads to better performance. A: distribution of median reaction times (RT) shown separately for Switch and Repetition trials. Light gray lines represent individual participants. Dashed black line shows group mean across conditions, with relative SE. B: mean RT as a function of repetition number (1 means that a sequence was performed twice in a row, 3+ means the average of a sequence being repeated 4 or more times in a row). Shaded areas represent between-subject SE. C: distribution of median sequence movement times (MT) shown separately for Switch and Repetition trials. Other conventions are the same as in A. D: mean sequence MT as a function of repetition number. Other conventions are the same as in B. *P < 0.05, two-tailed paired-samples t test.

Fig. 3.

Sequence repetition benefits online planning of repeating sequence elements. A: mean interpress intervals (IPI) as a function of transition number, shown separately for Switch (light gray) and Repetition (dark gray) trials. Shaded areas represent between-subject SE. B: same data as in A but normalized by mean IPI for each transition separately. Dashed gray line denotes group mean across conditions, with relative SE. *P < 0.05, two-tailed paired-samples t test.

Fig. 4.

The repetition effect appears to require movement experience. A: distribution of median movement times (MT) shown separately for Switch (light gray) and Repetition (dark gray) trials as a function of whether the previous (N − 1) trial was a Go or a No-Go trial. To avoid contamination between Go and No-Go trials in long repetition chains, selected trials were restricted to a maximum of one repetition. Solid light gray lines represent individual participants. Dashed black lines represent group means across conditions. Black dots are considered group outliers. B: distribution of median reaction times (RT) shown separately for Switch (light gray) and Repetition (dark gray) trials as a function of whether the previous (N − 1) trial was a Go or a No-Go trial. Other conventions are the same as in A. C: mean interpress intervals (IPI) as a function of transition number, shown separately for Switch (light) and Repetition (dark) trials and split by whether the preceding trial was a No-Go (left) or a Go (right) trial. Shaded areas represent between-subject SE. *P < 0.05, two-tailed paired-samples t test; ^•P < 0.05, interaction in 2-by-2 within-subject repeated measures ANOVA.

Fig. 5.

The repetition benefit remains constant across a wide range of movement speeds. A: sequence movement time (MT) as a function of MT percentile (11 bins) shown separately for Switch (light gray) and Repetition (dark gray) trials and by whether previous trial was No-Go (left) or Go (right). Dashed vertical line denotes the MT bin that includes the median MT. Shaded areas represent between-subject SE. B: repetition difference normalized by median MT for each bin, shown separately as a function of MT percentile and by whether previous trial was No-Go (light gray) or Go (dark gray). Other conventions are the same as in A. C: repetition difference normalized by median MT, plotted against MT for each participant. Solid line represents linear regression. Dotted line is a landmark for lack of repetition difference. R², proportion of variance in repetition difference explained by execution speed.

Fig. 6.

Sequence-specific learning reduces the repetition effect. A: sequence movement time (MT) shown separately for Switch (light gray) and Repetition (dark grays) trials as a function of block number in experiment (Exp.) 1 (left) and Exp. 2 (right). Shaded areas represent between-subject SE. B: Switch-Repetition difference normalized as a percentage of MT for each block in Exp. 1 (left) and Exp. 2 (right). Shaded areas represent between-subject SE. Solid black line represents linear regression line. Dashed horizontal line indicates absence of repetition effect. *P < 0.05, two-tailed paired-samples t test; ^••P < 0.05, two-tailed one-sample t test vs. zero difference; n.s., not significant.

Fig. 7.

Conceptual visualization of planning and execution in a neural state-space framework. Changes in the pattern of neuronal firing can be characterized by movements of the neuronal state in a low-dimensional state-space. During the preparation phase, these changes mostly occur in the planning state-space (bottom plane), whereas during the movement phase, these changes are more pronounced in the execution state-space (top plane). Different finger movements (corresponding to numbers; see Fig. 1) are characterized by a unique pattern in the planning state-space and a unique trajectory in the execution state-space. Dotted lines indicate temporal correspondence between state-space events across planning and execution planes. A: individual finger movements. During preplanning (after stimulus onset), the neural state trajectory moves from a baseline location (cross) toward the optimal planning state (gray spotlight) until the go signal (dot) triggers execution dynamics. On Repetition trials, the correct planning state would be reached more quickly and with higher accuracy, enabling faster movement initiation when the go signal is given. B: sequence of finger movements. Note that in this scenario, what improves upon repetition is the neural state trajectory on the planning plane (i.e., online planning), leaving neural dynamics unchanged on the execution plane.