Untitled Document

Review: Measuring Learning Progress via Self-Explanations versus Problem Solving-A
Suggestion for Optimizing Adaptation in Intelligent Tutoring Systems

By: - Christine Otieno, Rolf Schwonke, Alexander Renkl, Vincent Aleven, Ron Salden
Presented at: - CogSci 2011

The paper discusses about the best probable method for better learning in students. The idea, used in Cognitive tutors (Intelligent tutoring systems), could help improve the teaching- learning systems.
The experiment described in the paper discusses the relative advantage of problem solving, self-explanations and combined effect of both in learning processes. The subjects were divided in three groups.
GROUP 1: students were given worked out examples in predefined format of learning (fixed fading condition).
GROUP 2: examples were given individually in accordance with the skill level and self explanation performance (adaptive fading condition).
GROUP 3: base set of students(problem condition).
Problem set consisted of 15 mathematical problems involving 4 mathematical concepts (1st eight being from 1 concept). To solve a problem a subject had to write the answer and the reason.
The test included a pretest and a posttest (along with a delayed posttest).
Markers :
Prior specific knowledge by pretest.
Learning and self explanation by reason part of problems.
Inferences

Posttest scores (both immediate and delayed) showed greatly on reason part (self explanation) with no significant contribution in answer part.
Both problem solving and self explanation aid in the learning process.
The problem solving steps and prior knowledge enhances learning but this is overpowered by self explanation approaches.

Notes :
One may argue that problem solving along with self explanation is beyond the cognitive capacity in early stages of acquiring skills. But this may be neutralized by scaffolding effect by worked out examples. The cognitive tutor can be improved by inculcating example based learning and involving personal perspectives such as fading (Fading is a reduction in assistance over time).
References:

Preparing Students for Effective Explaining of Worked Examples in the Genetics Cognitive Tutor by Albert Corbett, Ben MacLaren, Angela Wagner, Linda Kauffman, Aaron Mitchell
Cognitive Load Theory: A Special Issue of Educational Psychologist By Fred Paas, Alexander Renkl, John Sweller

Worked Examples and the Assistance Dilemma by Ron Salden Vincent Aleven Rolf Schwonke
The Expertise Reversal Effect bySlava Kalyuga, Paul Ayres, Paul Chandler & John Sweller