The recommendation for learners to be provided with control over their feedback schedule is questioned in a self-controlled learning paradigm

ABSTRACT

Evidence that self-controlled feedback schedules are more effective for motor learning than yoked or predetermined schedules has been used to forward the recommendation that practitioners should provide choice to learners over when they would like to receive feedback. This recommendation can be questioned because the typical comparison groups in such experimentation do not well represent the applied setting. Consequently, comparison groups that better map onto the applied setting are needed. To this end, three groups learned a golf putting task: (1) self-controlled, (2) traditional-yoked, and (3) a group who were led to believe their KR schedule was being controlled by a golf coach. Participants (N = 60) completed a pre-test, acquisition phase, and delayed post-tests (retention/transfer). No group differences during the post-tests for mean radial error, F(2, 54) = 2.71, p = .075, or bivariate variable error, F(2, 56) = 0.11, p = .896, were found. Thus, the typical self-controlled learning advantage was not observed. Given the failure to replicate self-controlled benefits, combined with the fact there is little research that has directly compared self-controlled feedback schedules to coach-controlled schedules, we argue more evidence is needed before advocating that learners be provided choice over their feedback schedule.

KEYWORDS:

• Motor learning
• self-controlled feedback
• autonomy
• coach-control
• knowledge-of-results

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ One participant was excluded from data collection and a subsequent participant was recruited after finding out after the Day 1 procedure that they were still suffering from concussion-like symptoms.

² A putting experience was defined as anything from one round of golf on a golf course (either a standard 18-hole course or a 9-hole course) to one round of miniature putt.

³ During the pre-test, 17.2% of trials hit the backstop and were therefore trimmed to be 90 cm long, this percentage was lower for all other phases of the experiment with 3.9% in acquisition, 8.3% in retention, and 2.8% in transfer. Additionally, 0.5% of putts were shorter than 140 cm during the pre-test phase meaning they were not visible for the camera to get an accurate measurement, therefore these putts were trimmed and given a y-axis score of −140 cm. During acquisition, 0.1% of putts were less than −140 cm, followed by 0% in retention and 0.2% in transfer.

⁴ A secondary analysis was conducted for both dependent variables that compared pre-test to retention test performance to determine if there was learning in this experiment. Both one-way ANOVAs for MRE and BVE were significant, F(1, 58) = 10.86, p = .002, and F(1, 58) = 4.03, p = .049.

⁵ A sensitivity analysis was conducted leaving the two univariate outliers in the sample and this analysis did not change the significance of the results. Additionally, we conducted a traditional 3 (Group) x 2 (Test) ANOVA as another sensitivity analysis, and this also did not change the outcome of the results as no significant main effect for Group was found.

⁶ We also conducted a traditional 3 (Group) x 2 (Test) ANOVA for BVE as a sensitivity analysis and this revealed the same pattern of results with no significant differences obtained.

⁷ Sensitivity analyses were conducted with these four participants excluded and the significance did not change for any main effect or interaction for all MRE and BVE analyses.

⁸ These exploratory analyses were conducted due to reviewer comments.

⁹ Exclusion of outliers did not meaningfully change the results.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.