Supporting (aspects of) self-directed learning with Cognitive Tutors

Ken KoedingerCMU Director of Pittsburgh Science of Learning Center

Human-Computer Interaction & Psychology

Carnegie Mellon University

Tutoring Self-Directed Learning• Cognitive Tutors• Extending Cognitive Tutors to support

meta-cognitive processes:– Tutoring self-explanation– Tutoring help-seeking

Real World Impact of Cognitive & Learning Science

Cognitive Tutor Algebra• Combines Cognitive Psych & Artificial

Intelligence -> computational models of student thinking & learning

• Full course used in 2000 schools!

• Company: Carnegie Learning

Koedinger, Anderson, Hadley, & Mark (1997). Intelligent tutoring goes to school in the big city.

3(2x + 5) = 9

6x + 15 = 9 2x + 5 = 3 6x + 5 = 9

Cognitive Tutor Technology:Use ACT-R theory to individualize instruction

• Cognitive Model: A system that can solve problems in the various ways students can

If goal is solve a(bx+c) = dThen rewrite as abx + ac = d

If goal is solve a(bx+c) = dThen rewrite as abx + c = d

If goal is solve a(bx+c) = dThen rewrite as bx+c = d/a

• Model Tracing: Follows student through their individual approach to a problem -> context-sensitive instruction

• Cognitive Model: A system that can solve problems in the various ways students can

3(2x + 5) = 9

6x + 15 = 9 2x + 5 = 3 6x + 5 = 9

Cognitive Tutor Technology:Use ACT-R theory to individualize instruction

If goal is solve a(bx+c) = dThen rewrite as abx + ac = d

If goal is solve a(bx+c) = dThen rewrite as abx + c = d

• Model Tracing: Follows student through their individual approach to a problem ->context-sensitive instruction

Hint message: “Distribute a across the parentheses.”

Bug message: “You need tomultiply c by a also.”

• Knowledge Tracing: Assesses student's knowledge growth -> individualized activity selection and pacing

Known? = 85% chance Known? = 45%

Cognitive Tutors as Research PlatformPast Success:• Cognitive Tutors as delivery vehicle

– Bring existing Learning Science to classroom

New Goal:• Cognitive Tutors as research platform

– Create new Learning Science & Technology

• 5 year, $25 million research center:

Tutoring Self-Directed Learning• Extending Cognitive Tutors to support

meta-cognitive processes:– Tutoring self-explanation– Tutoring help-seeking– Tutoring error self-detection & correction

• See Tuesday paper: Mathan & Koedinger

Others:– Tutoring scientific inquiry– Tutoring good collaboration– Tutoring peer-to-peer tutoring

Tutoring Self-Directed Learning 1: Encourage Active Declarative Processing Through Self-Explanation

Aleven, V. & Koedinger, K. R. (2002). An effective metacognitive strategy: Learning by doing and explaining with a computer-based Cognitive Tutor. Cognitive Science, 26(2)

Problem: Shallow knowledge acquisition• Variations on shallow knowledge

– Over-general procedural knowledge • right for wrong reason

– No declarative k -- cannot explain, transfer

• Geometry example– “Looks-equal” production rule – If the goal is to find angle A

and it looks equal to angle B and angle B is D degreesThen conclude that angle A is D degrees

Example of Shallow Reasoning

Explanation Condition

Problem solving answers

Explanation by reference

Problem Solving Condition

Assessing transfer: “Not Enough Info” item

Assessing transfer: Incorrect over-generalization

SE Study 2 Results

Answer Items Reason Items Not Enough Info Items

.2

.3

.4

.5

.6

.7

.8

Proportion Correct on Post-Test

Reason

Answer Only

Numerical steps Explanations Transfer items

Problem Solving

Explanation

Condition

Extra Practice in Problem Solving => More Shallow Learning

Easy to guessitems

Hard to guessitems

.3

.4

.5

.6

.7

.8

.9

1

% C

orr

ect

Explanation

Problem Solving

Condition

Tutoring Self-Explanation Summary

• When Ss explain they learn more & learn with greater understanding:– better explanations of answers– better on harder-to-guess test items– better on transfer questions

• Possible to achieve benefits of self-explanation with simple manipulation

Tutoring Self-Directed Learning 2: Independent help-seeking skills

Tutoring Help-Seeking

• Goal: Foster long-term learner independence

• Model of ideal learning & help-seeking behaviors

QuickTime™ and aGraphics decompressor

are needed to see this picture.

• Tutor this model• Improve robust learning

– Long-term retention– Transfer– Accelerated future learning

Initial Classroom StudyHelp Tutor: 1) Detects meta-cognitive errors 2) These are correlated with poor learning

Help-seeking errors overall

General errors

Help avoidance

Help abuse

Try-step avoidance

Try-step abuse

Error rate 17% 1% 5% 6% 6% <0.5%

Correlation with

learning-0.42** -0.34* -0.41** -0.17 -0.27 -0.1

Measuring Future Learning

QuickTime™ and aTIFF (LZW) decompressor

are needed to see this picture.

Summary

• Cognitive Tutors can support student learning of aspects of better self-directed learning strategies

• More research: Inquiry, collaboration, …• Cognitive Tutors & LearnLab provide an

ideal platform to perform such research– See LearnLab.org

Related Research Efforts

• Pittsburgh Science of Learning Center

• Tools for authoring Cognitive Tutors– Variety of new domains: sciences,

languages

• “Assistments” for on-line dynamic assessment

ctat.pact.cs.cmu.edu

assistment.org

learnlab.org

END