Difference between revisions of "Roll help seeking principle"
(→Experimental support) |
(→Experimental support) |
||
| Line 19: | Line 19: | ||
==Experimental support== | ==Experimental support== | ||
| − | + | ||
===Laboratory experiment support=== | ===Laboratory experiment support=== | ||
| + | |||
| + | None that we know of. | ||
| + | |||
===In vivo experiment support=== | ===In vivo experiment support=== | ||
| − | Our current results show that meta-cognitive tutoring yields better meta-cognitive behavior within the tutored environment, but fail to generate a persistent effect (i.e., better meta-cognitive behavior in a novel context) or better domain learning (as evaluated using pre- and post-tests). | + | There is much support for supporting metacognition and reaching an effect within the tutored environment. Most notably, the work on self-explanation supports that (e.g., Aleven and Koedinger 2002; Conati and VanLehn 1999). |
| + | Also, it is known that systems with hints are better than no hints, and that contingent hints are better than conventional hints. (Wood & Wood, 1999) | ||
| + | However, these systems support meta-cognition within the scope of the system and do not thrive to affect behavior in a persistent matter, outside the tutoring environment. | ||
| + | |||
| + | So far we were not able to confirm (or dispute) this principle. Our current results show that meta-cognitive tutoring yields better meta-cognitive behavior within the tutored environment, but fail to generate a persistent effect (i.e., better meta-cognitive behavior in a novel context) or better domain learning (as evaluated using pre- and post-tests). Please see our [[project page|The Help Tutor Roll Aleven McLaren]] for more details. | ||
==Theoretical rationale== | ==Theoretical rationale== | ||
Revision as of 23:48, 13 January 2008
Contents
Brief statement of principle
Support metacognition in the context of problem solving using established principles of cognitive tutoring.
Description of principle
Instruction should yield a persistent effect on students meta-cognitive skills in a fashion that persists beyond the scope of the tutored environment. That is, the tutored practice should help students become better at learning.
This is the common effect of tutoring at the cognitive level - for example, students remember Pythagorean theorem also outside the specific instructional unit. Therefore, perhaps meta-cognitive tutoring could use established principles of cognitive tutoring.
Operational definition
Meta-cognitive tutoring should:
- Giving direct instruction
- Giving immediate feedback on errors
- Prompting for self-assessment
Examples
Experimental support
Laboratory experiment support
None that we know of.
In vivo experiment support
There is much support for supporting metacognition and reaching an effect within the tutored environment. Most notably, the work on self-explanation supports that (e.g., Aleven and Koedinger 2002; Conati and VanLehn 1999). Also, it is known that systems with hints are better than no hints, and that contingent hints are better than conventional hints. (Wood & Wood, 1999) However, these systems support meta-cognition within the scope of the system and do not thrive to affect behavior in a persistent matter, outside the tutoring environment.
So far we were not able to confirm (or dispute) this principle. Our current results show that meta-cognitive tutoring yields better meta-cognitive behavior within the tutored environment, but fail to generate a persistent effect (i.e., better meta-cognitive behavior in a novel context) or better domain learning (as evaluated using pre- and post-tests). Please see our The Help Tutor Roll Aleven McLaren for more details.
Theoretical rationale
(These entries should link to one or more learning processes.)
