Transfer of training benefits requires rules we cannot see (or hear).

K. G. Garner,Casey R. Lynch,Paul E. Dux
DOI: https://doi.org/10.1037/xhp0000215
2016-08-01
Abstract:Although humans show a remarkable ability to make rapid and accurate decisions in novel situations, it is surprisingly difficult to observe transferable benefits when training decision-making performance. The current study investigated whether 2 properties of decision-making-amodal processing and encoding of abstract relationships-could be leveraged to produce transferable training gains, compared with the performance of an active-control group. Experiment 1 showed that training responses to visually presented stimuli (letters) did not transfer to benefit performance for the same stimuli presented in the auditory modality. Therefore, training exercises the integration of modality-specific information, not a supramodal category. However, Experiment 2 showed that when stimuli share an abstract rule, training transfers to new materials that conform to the same modality and rule, and to analogous rules in a new modality. Therefore, transfer of training benefits requires an abstract code that can be generalized to new stimulus sets. (PsycINFO Database Record
psychology, experimental
What problem does this paper attempt to address?