Limits...
Learning from the value of your mistakes: evidence for a risk-sensitive process in movement adaptation.

Trent MC, Ahmed AA - Front Comput Neurosci (2013)

Bottom Line: We found that adaptation indeed differed.Specifically, in the Unstable environment, we observed reduced adaptation to leftward errors, an appropriate strategy that reduced the chance of a penalizing rightward error.These results demonstrate that adaptation is influenced by the subjective value of error, rather than solely the magnitude of error, and therefore is risk-sensitive.

View Article: PubMed Central - PubMed

Affiliation: Neuromechanics Laboratory, Department of Integrative Physiology, University of Colorado, Boulder Boulder, CO, USA.

ABSTRACT
Risk frames nearly every decision we make. Yet, remarkably little is known about whether risk influences how we learn new movements. Risk-sensitivity can emerge when there is a distortion between the absolute magnitude (actual value) and how much an individual values (subjective value) a given outcome. In movement, this translates to the difference between a given movement error and its consequences. Surprisingly, how movement learning can be influenced by the consequences associated with an error is not well-understood. It is traditionally assumed that all errors are created equal, i.e., that adaptation is proportional to an error experienced. However, not all movement errors of a given magnitude have the same subjective value. Here we examined whether the subjective value of error influenced how participants adapted their control from movement to movement. Seated human participants grasped the handle of a force-generating robotic arm and made horizontal reaching movements in two novel dynamic environments that penalized errors of the same magnitude differently, changing the subjective value of the errors. We expected that adaptation in response to errors of the same magnitude would differ between these environments. In the first environment, Stable, errors were not penalized. In the second environment, Unstable, rightward errors were penalized with the threat of unstable, cliff-like forces. We found that adaptation indeed differed. Specifically, in the Unstable environment, we observed reduced adaptation to leftward errors, an appropriate strategy that reduced the chance of a penalizing rightward error. These results demonstrate that adaptation is influenced by the subjective value of error, rather than solely the magnitude of error, and therefore is risk-sensitive. In other words, we may not simply learn from our mistakes, we may also learn from the value of our mistakes.

No MeSH data available.


Related in: MedlinePlus

Sensitivity vs. Gain. Group averaged elements of the sensitivity vector are plotted for Stable (blue) and Unstable (red) phases. For clarity, arrows are used to indicate gains resulting in increasingly leftward or rightward errors. Asterisks indicate P < 0.05. Error bars represent standard error of the mean.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3750521&req=5

Figure 7: Sensitivity vs. Gain. Group averaged elements of the sensitivity vector are plotted for Stable (blue) and Unstable (red) phases. For clarity, arrows are used to indicate gains resulting in increasingly leftward or rightward errors. Asterisks indicate P < 0.05. Error bars represent standard error of the mean.

Mentions: We next turn to the model gain-based analysis of adaptation. While the elements of the sensitivity vector, S, varied from participant to participant, the values of D (2.2e-4 ± 0.46e-4 m2/Ns) and A (0.69 ± 0.08) are consistent with previous findings (Fine and Thoroughman, 2007). Model fits predicted the data well with an average R2 of 0.76 ± 0.002 across participants. The elements of the sensitivity vector exhibit a linear positive relationship with gain in both phases (Figure 7). Again these data were analyzed using the linear mixed effects regression model with gain and phase included as factors, as well as a gain by phase interaction term. The slope of the sensitivity vs. gain curve during both the Stable and Unstable phase were found to be significant: P < 0.002, k = 0.60519 and P < 0.002, k = 0.4188, respectively. As in the behavioral results, there was a significant gain by phase interaction (P = 0.0000168), a planned comparison at the largest gain indicates a significant difference (P = 0.0006; B = −40 Ns/m). Specifically, sensitivity to this gain was reduced in the Unstable phase compared to the Stable phase. Post-hoc pairwise comparisons also reveal a trend toward a reduction in sensitivity to gains B = −36 Ns/m, B = −24 Ns/m, and −20 Ns/m (P = 0.0222, P = 0.0107, P = 0.0435). However, when correcting for multiple comparisons these P-values are greater than the Bonferroni-corrected significance level (α = 0.05/9 = 0.0056). Overall, these model results confirm our behavioral findings. They indicate that sensitivity, the model-based metric of adaptation, was also affected by phase, and that sensitivity to the larger gains was reduced in the Unstable phase compared to the Stable phase. These findings are also in line with the predictions in Figure 3 in the case of reduced adaptation to leftward errors only (i.e., the strongest gains). Moreover, compared to the behavioral results, the sensitivity analysis reveals a stronger effect of phase on sensitivity. We observed reduced sensitivity in four of the highest gains, compared to the observation of reduced adaptation to only the largest gain in the behavioral analysis.


Learning from the value of your mistakes: evidence for a risk-sensitive process in movement adaptation.

Trent MC, Ahmed AA - Front Comput Neurosci (2013)

Sensitivity vs. Gain. Group averaged elements of the sensitivity vector are plotted for Stable (blue) and Unstable (red) phases. For clarity, arrows are used to indicate gains resulting in increasingly leftward or rightward errors. Asterisks indicate P < 0.05. Error bars represent standard error of the mean.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3750521&req=5

Figure 7: Sensitivity vs. Gain. Group averaged elements of the sensitivity vector are plotted for Stable (blue) and Unstable (red) phases. For clarity, arrows are used to indicate gains resulting in increasingly leftward or rightward errors. Asterisks indicate P < 0.05. Error bars represent standard error of the mean.
Mentions: We next turn to the model gain-based analysis of adaptation. While the elements of the sensitivity vector, S, varied from participant to participant, the values of D (2.2e-4 ± 0.46e-4 m2/Ns) and A (0.69 ± 0.08) are consistent with previous findings (Fine and Thoroughman, 2007). Model fits predicted the data well with an average R2 of 0.76 ± 0.002 across participants. The elements of the sensitivity vector exhibit a linear positive relationship with gain in both phases (Figure 7). Again these data were analyzed using the linear mixed effects regression model with gain and phase included as factors, as well as a gain by phase interaction term. The slope of the sensitivity vs. gain curve during both the Stable and Unstable phase were found to be significant: P < 0.002, k = 0.60519 and P < 0.002, k = 0.4188, respectively. As in the behavioral results, there was a significant gain by phase interaction (P = 0.0000168), a planned comparison at the largest gain indicates a significant difference (P = 0.0006; B = −40 Ns/m). Specifically, sensitivity to this gain was reduced in the Unstable phase compared to the Stable phase. Post-hoc pairwise comparisons also reveal a trend toward a reduction in sensitivity to gains B = −36 Ns/m, B = −24 Ns/m, and −20 Ns/m (P = 0.0222, P = 0.0107, P = 0.0435). However, when correcting for multiple comparisons these P-values are greater than the Bonferroni-corrected significance level (α = 0.05/9 = 0.0056). Overall, these model results confirm our behavioral findings. They indicate that sensitivity, the model-based metric of adaptation, was also affected by phase, and that sensitivity to the larger gains was reduced in the Unstable phase compared to the Stable phase. These findings are also in line with the predictions in Figure 3 in the case of reduced adaptation to leftward errors only (i.e., the strongest gains). Moreover, compared to the behavioral results, the sensitivity analysis reveals a stronger effect of phase on sensitivity. We observed reduced sensitivity in four of the highest gains, compared to the observation of reduced adaptation to only the largest gain in the behavioral analysis.

Bottom Line: We found that adaptation indeed differed.Specifically, in the Unstable environment, we observed reduced adaptation to leftward errors, an appropriate strategy that reduced the chance of a penalizing rightward error.These results demonstrate that adaptation is influenced by the subjective value of error, rather than solely the magnitude of error, and therefore is risk-sensitive.

View Article: PubMed Central - PubMed

Affiliation: Neuromechanics Laboratory, Department of Integrative Physiology, University of Colorado, Boulder Boulder, CO, USA.

ABSTRACT
Risk frames nearly every decision we make. Yet, remarkably little is known about whether risk influences how we learn new movements. Risk-sensitivity can emerge when there is a distortion between the absolute magnitude (actual value) and how much an individual values (subjective value) a given outcome. In movement, this translates to the difference between a given movement error and its consequences. Surprisingly, how movement learning can be influenced by the consequences associated with an error is not well-understood. It is traditionally assumed that all errors are created equal, i.e., that adaptation is proportional to an error experienced. However, not all movement errors of a given magnitude have the same subjective value. Here we examined whether the subjective value of error influenced how participants adapted their control from movement to movement. Seated human participants grasped the handle of a force-generating robotic arm and made horizontal reaching movements in two novel dynamic environments that penalized errors of the same magnitude differently, changing the subjective value of the errors. We expected that adaptation in response to errors of the same magnitude would differ between these environments. In the first environment, Stable, errors were not penalized. In the second environment, Unstable, rightward errors were penalized with the threat of unstable, cliff-like forces. We found that adaptation indeed differed. Specifically, in the Unstable environment, we observed reduced adaptation to leftward errors, an appropriate strategy that reduced the chance of a penalizing rightward error. These results demonstrate that adaptation is influenced by the subjective value of error, rather than solely the magnitude of error, and therefore is risk-sensitive. In other words, we may not simply learn from our mistakes, we may also learn from the value of our mistakes.

No MeSH data available.


Related in: MedlinePlus