Prolonged Feedback Duration Does Not Affect Implicit Recalibration in a Visuomotor Rotation Task

Abstract

Visuomotor rotations are frequently used to study cognitive processes underlying motor adaptation. Explicit aiming strategies and implicit recalibration are two of these processes. A large body of literature indicates that both processes are in fact dissociable and mainly independent components that can be measured using different manipulations in visuomotor rotation tasks. Visual feedback is a crucial element in these tasks, and it therefore plays an important role when assessing explicit re-aiming and implicit recalibration. For instance, researchers have found timing of visual feedback to affect the contribution of implicit recalibration to learning: if feedback is shown only at the end of the movement (instead of continuously), implicit recalibration decreases. Similarly, participants show lower levels of implicit recalibration if visual feedback is presented with a delay (instead of immediately). We thus hypothesized that the duration of feedback availability might also play a role. The goal of this study was thus to investigate the effect of longer versus shorter feedback durations on implicit recalibration in human participants. To this end, we compared three feedback durations in a between-subject design: 200, 600, and 1200 ms. Using a large sample size, we found differences between groups to be quite small, to the point where most differences indicated statistical equivalence between group means. We therefore hypothesize that feedback duration, when only endpoint feedback is presented, has a negligible effect on implicit recalibration. We propose that future research investigate the effect of feedback duration on other parameters of adaptation, so as proprioceptive recalibration and explicit re-aiming.

Keywords: implicit recalibration; motor adaptation; visuomotor rotation.

Figure 1.

Experimental setup and design. A, The experiment is made up of four different phases: baseline, rotation, aftereffect, and washout. B, Trial structure. Each trial started with the participants moving their hand into the starting position (filled gray dot), where they stayed for 1000 ms (origin hold) until the target appeared (filled red dot). They had 500 ms to execute the movement and received visual feedback for 200, 600, or 1200 ms depending on their assigned group (shown as filled green dot). Visual feedback could be either rotated (Rotation) or not present (Aftereffect).

Figure 2.

Simplified diagram of the hierarchical Bayesian model showing how the key coefficients for our statistical comparison were determined.

Figure 3.

HDI widths as a function of sample size. Number of subjects refers to the number of subjects per group. The black line denotes the 90th percentile of the 10 model repetitions. The left figure shows the HDI width for each group mean, the right figure shows the HDI width for the group differences.

Figure 4.

Learning curve and aftereffect. A, Binned learning curve for all three groups (30 bins of eight trials each). Shaded area denotes 95% HDI. B, Mean movement direction during early and late baseline for all three groups. C, Mean movement direction during early and late aftereffect for all three groups. Error bars show 95% HDI for each group, vertical lines show 95% HDI for individual subjects.

Figure 5.

Differences between groups during early and late aftereffect. ΔSD,LD denotes the difference between the short duration and the long duration group, ΔSD,MD shows the difference between short duration and medium duration and ΔMD,LD the difference between the medium and the long duration group. Δ[°] is the difference in degrees. Dashed horizontal lines represent the ROPE of −3% through 3% around a value of 0 difference between the groups. Error bars show 95% HDI for the group differences.