Abstract
Two empirical effects, previously shown to be advantageous to learning, were contrasted in this study: a) self-regulation of practice trial order (vs. experimenter-defined), and b) retroactively-presented model information (vs. proactively-presented). The purpose was to assess if self-regulated practice modified the effects of the timing of modeled information for the learning of three sequential-timing patterns. Each of the patterns consisted of unique temporal parameters between the presses (and accompanying auditory tone) of the same sequence of five keys. Auditory information was modeled either before (proactively) or after (retroactively) each trial, and practice order was learner-regulated. Absolute constant error (|CE|) and variable error (VE) were calculated for both the total time (TT) and the sum of the segmental times (SST). Retention tests revealed that the proactive group performed with less |CE| for SST, while the retroactive group performed with less |CE| for TT for at least one pattern; suggesting that self-regulated practice eliminated the clear advantage for providing modeled information retroactively. The retention results also revealed that learners who chose to switch patterns frequently during practice, regardless of when the model was presented, had less |CE| and VE for SST and less VE for TT than learners who self-regulated with infrequent task switches; suggesting that learner-imposed contextual interference enhanced retention.Acknowledgments: This study was supported by NSERC