Difference between revisions of "Transfer of learning"

From Learning
Jump to: navigation, search
 
Line 28: Line 28:
  
 
The general ability referenced above might be contingent on learning in some other context ''C'', in which case we would say that there is significant transfer of learning from ''C'' to both ''A'' and ''B'' separately, but none from ''A'' to B''.
 
The general ability referenced above might be contingent on learning in some other context ''C'', in which case we would say that there is significant transfer of learning from ''C'' to both ''A'' and ''B'' separately, but none from ''A'' to B''.
 +
 +
==Directionality of transfer of learning==
 +
 +
Research suggests that transfer of learning is directional, i.e., it is asymmetric. There may be significantly greater transfer of learning from context ''A'' to context ''B'' than from context ''B'' to context ''A''. One key factor affecting directionality is the high road versus low road distinction. Transfer of learning that relies on the low road is easier, and within high road transfer, transfer that relies on backward reaching approaches is easier. For instance:
 +
 +
* There is more transfer of learning from the learning context of mathematical methods to the transfer context of another discipline where that mathematical method is used than vice versa. For instance, a study showed that interdomain transfer from the algebra of arithmetic progression to the (structurally isomorphic) physics of constant acceleration than in the other direction.
 +
 +
Note that this evidence of directionality is in partial contradiction with the [[examples first]] pedagogical idea where examples are presented first and general principles are then inferred from those. In fact, ''examples first'' is not pedagogically efficient from this perspective, but it may have other offsetting advantages.

Latest revision as of 19:42, 28 December 2013

Definition

Transfer of learning refers to a situation where learning (in the sense of knowledge, skills, or modes of thought) acquired in one context (which we call the learning context) are applied in another (ostensibly different) context (which we call the transfer context).

The extent to which transfer of learning occurs is a major determinant of the value of learning something.

Taxonomy of forms of transfer of learning

Classification Relevant criterion More on the terms
near versus far the extent of overlap between the learning. Near denotes significant overlap, far denotes negligible overlap.
positive versus negative whether the learning in the learning context helps or hinders performance in the transfer context. Positive indicates that it helps, negative indicates that it hinders.
vertical versus horizontal the extent to which the transfer context builds on, or depends upon, the learning context. Vertical indicates that the transfer context builds on the learning context, whereas horizontal suggests that there are similarities between the learning context and transfer context, but no dependence in either direction.
literal versus figural the extent to which the knowledge being used needs to be modified or customized to the new context. Literal indicates that intact knowledge transfers to the new task, figural indicates that the learning context helps with general skills to help tackle the transfer context.
low road versus high road the extent of automaticity with which the knowledge is transfered. Low road indicates that the knowledge is transfered automatically and subconsciously, high road indicates that connections need to be consciously formulated for successful transfer.
(within high road): forward reaching versus backward reaching the direction in which one reasons between the learning context and the transfer context. Forward reaching means abstracting the situation from the learning context to a potential transfer context, backward reaching means abstracting in the transfer context features of a previous situation where new skills and knowledge were learned.

Distinction between transfer of learning and existence of general abilities

Transfer of learning is closely related to the question of whether there exist general abilities to learn or to understand material. However, it is different, in that what we are specifically interested in is whether the learning, i.e., the change in knowledge or skills, transfers from one context to the other. It is possible, for instance, to take the view that performance in two contexts is correlated (i.e., that people who do better in context A also do better in context B) while still believing that there is little transfer of learning between the contexts, i.e., that becoming better at context A does not automatically make one better at context B. A plausible explanation for such a phenomenon is that performance in context A results from a combination of general (pre-existing) abilities and context-specific learning, and the correlation between context A and context B arises entirely from the general abilities part, so that improving performance in context A through context-specific learning does not improve performance in context B.

The general ability referenced above might be contingent on learning in some other context C, in which case we would say that there is significant transfer of learning from C to both A and B separately, but none from A to B.

Directionality of transfer of learning

Research suggests that transfer of learning is directional, i.e., it is asymmetric. There may be significantly greater transfer of learning from context A to context B than from context B to context A. One key factor affecting directionality is the high road versus low road distinction. Transfer of learning that relies on the low road is easier, and within high road transfer, transfer that relies on backward reaching approaches is easier. For instance:

  • There is more transfer of learning from the learning context of mathematical methods to the transfer context of another discipline where that mathematical method is used than vice versa. For instance, a study showed that interdomain transfer from the algebra of arithmetic progression to the (structurally isomorphic) physics of constant acceleration than in the other direction.

Note that this evidence of directionality is in partial contradiction with the examples first pedagogical idea where examples are presented first and general principles are then inferred from those. In fact, examples first is not pedagogically efficient from this perspective, but it may have other offsetting advantages.