Sanchez, D. J., Yarnik, E. N., & Reber, P. J. (submitted). Specificity of transfer in perceptual-motor sequence learning.
Examining the transfer of skilled knowledge can tell us about both the underlying representation of what has been learned and how this information is applied in contexts that differ from training. A common model of transfer is that performance in the new context will be based on the overlap of component processes between training and test conditions. However, we recently reported (Gobel, Sanchez & Reber, 2011) that a selective change to one aspect of performance (timing) apparently disrupted application of any learned information in an implicit skill learning task. This implies an important role for integration of information that limits transfer across context. Two experiments tested the degree to which integration of information across separable information components limited transfer from training to test using a perceptual-motor sequence learning task. In Experiment 1, selective disruption of timing or order information was re-examined using a novel response manipulandum that allowed for analysis of each response component and more sensitive detection of transfer. In Experiment 2, transfer was examined after selective disruption of perceptual information that left the motor response sequence intact. Both experiments found evidence of partial transfer that was less than would be predicted by simple overlap of information from training to test. The relatively weak transfer implies that implicit skill learning also occurs in the integration of information, even across modalities, meaning that learning will generally be specific to training context and full transfer of knowledge will be difficult to achieve.