Sequence learning in the motor domain to date has only been demonstrated in combination with visual learning. The present study sought to isolate motor and visual sequence learning.

To this end, the response time (RT) in a button-pressing task was measured. A coloured circle in the screen center served as the target, and four coloured arcs around it mapped colour to four buttons’ locations. The arcs’ colours were changeable to uncouple the visual and the motor sequence. In the visual learning condition, target colours were sequential, while response buttons were random, and vice versa in the motor condition. Sequential lists were used for the training blocks, and randomized lists were used to measure baseline performance. Participants (n=30, mean age=24.2, 12 male) were randomly assigned to two groups: V1M2 (visual first day, motor second day) or M1V2 (motor first day, visual second day).

A mixed 2 (block: sequential vs random) * 2 (condition: visual vs motor) * 2 (group: V1M2 vs M1V2) ANOVA for the RT showed a main effect of the block, F(1,28)=32.333, p<.001, η_p²=0.536. Sequential RT was shorter (561 ms) than random RT (571 ms), indicating that sequence learning occurred. There were no interactions of block*condition, F(1,28)=0.020, p=.888, η_p²=0.001, or of block*condition*group, F(1,28)=0.241, p=.627, η​​​​​​​_p²=0.009, indicating that sequence learning did not differ between the visual and motor condition or between groups. More importantly, these results show that sequence learning in the motor domain can occur completely independent of visual learning.