Task integration and anticipation in complex, continuous motor tasks

Abstract

Multitasking and sequential motor learning research has advanced greatly in recent years, yet commonly accepted insights are largely based on simple, distinct tasks which cannot accurately reflect the variety of more complex and continuous tasks we encounter in everyday life. This study therefore aims to reassess the influence of task integration on motor sequence learning in complex, continuous tasks through the use of a virtual reality environment and an adapted SRT dual task suited for continuous movements. In our experiment, participants performed a complex, bimanual motor sequence task with varying degrees of suitability for task integration. We could successfully show that task integration has beneficial effects on complex task acquisition if covariations between tasks are consistent and detrimental effects if covariations are too inconsitent or missing. Minor inconsistencies within a repeated sequence can however be mitigated. These results highlight the distinct influence of task integration on complex, continuous motor learning, yet emphasize the need for further research beyond distinct, simple tasks.

Keywords: SRT task; anticipation; complex motor task; implicit motor learning; sequence learning; task integration.

Figure 1

Task structure for dual-task groups (n = 20). The repeating right-hand sequence shared by all groups is represented by the letters “A–F,” numbers represent repeating sequence elements and “r” random stimuli. For each non-catch-trial block, the sequence was repeated eight times for a total of 72 trial pairs. The Parallel group continued the next sequence with the successor of the stimulus last completed, here with stimulus No. 2.

Figure 2

Main task as seen from participants’ perspective. Cubes appeared towards the far side of the indicated path before moving towards participants on a straight trajectory. A combo counter on the left, which added the number of hits without mistakes, and a points multiplier on the right served as feedback for participants’ performance.

Figure 3

Performance changes during practice phase. The graphs show the means (95% CI) of each practice block for each variable and group. All groups significantly improved their performance over the course of the practice phase. CB improved notably after the break between blocks six and seven, while DtC and TD improved more gradually. No significant differences between groups were found.

Figure 4

Performance changes during test phase. The graphs show the means (95% CI) of each test-phase block for each variable and group. The catch trial in blocks three and four lead to significant performance detriments, except for the Random, Partial 1, and Partial 2 group for DtC, and the Parallel group for TD.

Figure 5

Implicit learning scores between groups. The respective implicit learning scores (mean 95% CI) for each group and each variable are shown. Fisher’s or Welch’s one-way between-groups ANOVAs with the dependent variable “implicit learning score” and the grouping variable “group” were used. Error bars represent standard errors. The Integrated and Partial 1 groups display the highest implicit learning scores for CB and TD, while the DtC scores are similar for all groups.