Assistance

From LearnLab
Revision as of 01:16, 23 May 2008 by Koedinger (talk | contribs)
Jump to: navigation, search

Assistance is a property of instruction. Intuitively, assistance refers to how much help the learner gets from the instruction as they are learning. The term scaffolding is nearly synonymous.

Operationally, assistance is equated with the probability of a good performance during training. For instance, if the training involves solving a multi-step problem, and the experimental instruction causes students to have a higher probability of correctly entering steps than the control instruction, then the experimental instruction has a higher degree of assistance.

Theoretically, one can decide whether one instructional method provides more assistance than another by analysis of whether one gives more information than the other. For instance, a worked example gives more information than the analogous problem because the worked example gives the answer and the steps to solve the problem whereas the problem withholds the answer and steps. In other words, because a problem provides a logical subset of the information provided by a matched worked example, a problem is less assistance than a worked example.

It is important to emphasize that saying students get more instructional assistance from one instructional method than another does not mean students will learn more. Assistance is not equal to learning. Sometimes less assistance is better, for instance, because it may be more engaging or because it may require more deep thought or sense making. This difficult challenge for an instructional designer of deciding when to give more assistance or information versus when to withhold it is the assistance dilemma. That students with higher competence may benefit from lower levels of assistance is the assistance hypothesis.