Difference between revisions of "Roll help seeking principle"

From LearnLab
Jump to: navigation, search
(In vivo experiment support)
(Theoretical rationale)
Line 35: Line 35:
 
==Theoretical rationale==  
 
==Theoretical rationale==  
 
(These entries should link to one or more [[:Category:Learning Processes|learning processes]].)
 
(These entries should link to one or more [[:Category:Learning Processes|learning processes]].)
 +
 +
In short, Metacognitive tutoring is important since better metacognition knowledge should help students [[sense making | make sense]] of the material to be learned.
 +
Since metacognitive knowledge is type of knowledge, it should be doable to tutor it in a similar fashion to tutoring domain knowledge.
 +
 
==Conditions of application==
 
==Conditions of application==
 
==Caveats, limitations, open issues, or dissenting views==
 
==Caveats, limitations, open issues, or dissenting views==

Revision as of 00:02, 14 January 2008

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 project page for more details.

Theoretical rationale

(These entries should link to one or more learning processes.)

In short, Metacognitive tutoring is important since better metacognition knowledge should help students make sense of the material to be learned. Since metacognitive knowledge is type of knowledge, it should be doable to tutor it in a similar fashion to tutoring domain knowledge.

Conditions of application

Caveats, limitations, open issues, or dissenting views

Variations (descendants)

Generalizations (ascendants)

References