Individual differences in explicit and implicit visuomotor learning and working memory capacity

Abstract

The theoretical basis for the association between high working memory capacity (WMC) and enhanced visuomotor adaptation is unknown. Visuomotor adaptation involves interplay between explicit and implicit systems. We examined whether the positive association between adaptation and WMC is specific to the explicit component of adaptation. Experiment 1 replicated the positive correlation between WMC and adaptation, but revealed this was specific to the explicit component of adaptation, and apparently driven by a sub-group of participants who did not show any explicit adaptation in the correct direction. A negative correlation was observed between WMC and implicit learning. Experiments 2 and 3 showed that when the task restricted the development of an explicit strategy, high WMC was no longer associated with enhanced adaptation. This work reveals that the benefit of high WMC is specifically linked to an individual's capacity to use an explicit strategy. It also reveals an important contribution of individual differences in determining how adaptation is performed.

Figure 1. Spatial WM task.

Following a fixation period of 1 second, participants were asked to remember the positions of three, four, five or six red target circles presented simultaneously in a circular grid. Following a 3 second delay, a question mark (probe) appeared in or adjacent to one of the target positions. Participants were asked to make a button press with the index or middle finger of their right hand, to indicate whether a circle had appeared at the location indicated.

Figure 2. Visuomotor adaptation task.

(a) Experimental set up. Participants made reaching movements towards targets displayed on a screen. (b) Verbal report task (Taylor et al.10). The target was placed at 0 with numbers being placed either side of the target at 5° intervals. Participants were asked to report which number they were planning to move their finger towards. This was used as a direct measure of participant’s trial-by-trial cognitive strategy. (c) Experiment 1: participants adapted to an abrupt 45° visuomotor rotation with terminal feedback whilst performing the verbal report task (grey) prior to each movement. (d) Experiment 2: participants adapted to an abrupt 45° visuomotor rotation with terminal feedback. (e) Experiment 3: participants adapted to a gradual 45° visuomotor rotation with online feedback. The verbal report task (grey) was performed at the start and end of the adaptation block. Amount of trials and epoch number are provided in brackets. Between baseline, adaptation and washout there were short rest periods.

Figure 3. Experiment 1: explicit and implicit visuomotor adaptation.

(a,b) The significant correlation between WMC and hand direction (a) and between WMC and absolute hand direction error (b) during adaptation to an abrupt 45° visuomotor rotation. (c,d) The significant correlation between WMC and aiming direction, our measure of explicit adaptation derived from the verbal report task (c), and between absolute aiming direction error and WMC (d). (e,f) The significant correlation between WMC and hand direction minus aim direction (e), our measure of implicit adaptation, and between WMC and absolute values of hand direction minus aim direction (f). For illustration we show the angle that resulted from subtracting the angle of movement from the target angle (a,c,e), but for the statistical analyses we use absolute values representing the angle by which the implicit adaptation differed from 45° (b,d) or 0° (f). Using the absolute values as opposed to the raw values for the angles did not affect the results; the same correlations were statistically significant, irrespective of the measure of adaptation. The red rectangles show the data-points from the five participants who did not show any explicit adaptation in the direction of the target, and who drove the correlation. The dashed grey lines show that three of the participants who did not show any explicit adaptation in the direction of the target also showed high implicit adaptation.

Figure 4. Experiment 1: participants who did and did not show explicit adaptation in the direction of the target.

Hand direction group data (line = group mean, shaded area = standard error of mean across group) during adaptation to an abrupt 45° visuomotor rotation. The “High Exp” (n = 25, blue) group showed explicit adaptation (and aim angle of >0 in the direction of the target). The “Low Exp” (n = 5, red) group did not show any explicit adaptation in the direction of the target (they are the group within the red rectangles in Fig. 3). (a) The “High Exp” group showed greater adaptation overall. (b) Aiming direction (°) derived from the verbal report task for the two groups. (c) The “Low Exp” group showed greater implicit adaptation.

Figure 5. Experiment 2: Abrupt adaptation.

(a,b) The significant correlation between WMC and hand direction (a) and between WMC and absolute hand direction error (b) during adaptation to an abrupt 45° visuomotor rotation when participants did not report their hand direction. (c) For illustration we performed a median split, separating the group into those with high and low WMC. Hand direction throughout the experiment is shown for the high WMC (n = 17, blue) and low WMC (n = 17, red) groups (line = group mean, shaded area = standard error of mean across group). The high WMC group showed greater adaptation relative to the low WM group, however no differences were observed during baseline or washout.

Figure 6. Experiment 3: Gradual adaptation.

(a,b) Neither hand direction (a) not absolute hand direction error (b) during adaptation to a gradual 45° visuomotor rotation correlated with WMC. (c) Again, for illustration, hand direction is shown for the two groups (high and low WMC, based on a median split; line = group mean, shaded area = standard error of mean across group). The high WM (n = 10, blue) and low WM (n = 10, red) group showed similar adaptation. (d) No vision trials were used throughout to provide a relatively clean measure of sensorimotor recalibration (implicit process). There were no observable differences between high and low WM groups during adaptation but a suggestion that the low WM group showed a greater ‘after-effect’ during washout.