Learning Factors Analysis – A General Method for Cognitive Model Evaluation and Improvement Hao Cen, Kenneth Koedinger, Brian Junker Human-Computer Interaction.

Slides:



Advertisements
Similar presentations
Questions From Yesterday
Advertisements

High Resolution studies
AGVISE Laboratories %Zone or Grid Samples – Northwood laboratory
Using Graphics in the Math Classroom. Benefits to using the Graphics 1.Students become familiar with the common pictures they will see 2.Students learn.
Lesson 2.2. Students Connie and Oscar from Exercises 7 and 8 in Lesson 2.1 had the same mean and median scores, but the ranges and IQR for their scores.
STATISTICS Linear Statistical Models
Multiplication X 1 1 x 1 = 1 2 x 1 = 2 3 x 1 = 3 4 x 1 = 4 5 x 1 = 5 6 x 1 = 6 7 x 1 = 7 8 x 1 = 8 9 x 1 = 9 10 x 1 = x 1 = x 1 = 12 X 2 1.
1 When you see… Find the zeros You think…. 2 To find the zeros...
Jörg Drechsler (Institute for Employment Research, Germany) NTTS 2009 Brussels, 20. February 2009 Disclosure Control in Business Data Experiences with.
Jeopardy Topic 1Topic Q 1Q 6Q 11Q 16Q 21 Q 2Q 7Q 12Q 17Q 22 Q 3Q 8Q 13Q 18Q 23 Q 4Q 9Q 14Q 19Q 24 Q 5Q 10Q 15Q 20Q 25.
1/23 Learning from positive examples Main ideas and the particular case of CProgol4.2 Daniel Fredouille, CIG talk,11/2005.
Data Visualization Lecture 4 Two Dimensional Scalar Visualization
The National Certificate in Adult Numeracy
The 5S numbers game..
Solve Multi-step Equations
1 Lecture 2: Metrics to Evaluate Performance Topics: Benchmark suites, Performance equation, Summarizing performance with AM, GM, HM Video 1: Using AM.
The basics for simulations
Solving Multi-Step Equations
Factoring Quadratics — ax² + bx + c Topic
Chapter 4: Basic Estimation Techniques
1 Revisiting salary Acme Bank: Background A bank is facing a discrimination suit in which it is accused of paying its female employees.
Lengths of Complex Shapes
1 Generating Network Topologies That Obey Power LawsPalmer/Steffan Carnegie Mellon Generating Network Topologies That Obey Power Laws Christopher R. Palmer.
Areas of Complex Shapes
Area Of Shapes. 8cm 2cm 5cm 3cm A1 A2 16m 12m 10m 12cm 7cm.
Area in the amount of space inside an enclosed region. Area of Rectangle = base x height Base =10 Height = 6 Area = (10)(6) = 60 square units.
Advanced Methods and Analysis for the Learning and Social Sciences PSY505 Spring term, 2012 January 23, 2012.
MG Review.
Building Blocks 1 SG MATHEMATICS Credit. Qu. 1 If we write the number ABC DE in the form what is the value of n.
Solving One-Step Equations
Evaluation of an intervention to increase online filing of individuals’ tax returns Peter Lumb September 2009.
Statistical Analysis SC504/HS927 Spring Term 2008
Surface Area and Volume
More Two-Step Equations
Before Between After.
Benjamin Banneker Charter Academy of Technology Making AYP Benjamin Banneker Charter Academy of Technology Making AYP.
Slide R - 1 Copyright © 2009 Pearson Education, Inc. Publishing as Pearson Prentice Hall Active Learning Lecture Slides For use with Classroom Response.
Linear Relationships Vocabulary. Multiplying a number by -1 always gives you the opposite of that number -4 = 4 12 = -12 Multiplication Property of -1.
Linear Algebra (Mathematics for CG) Reading: HB Appendix A
Foundation Stage Results CLL (6 or above) 79% 73.5%79.4%86.5% M (6 or above) 91%99%97%99% PSE (6 or above) 96%84%100%91.2%97.3% CLL.
Non-linear Components for Multiple Regression Why we might need non-linear components Type of non-linear components Squared-term components & their disadvantages.
CSE 473/573 Computer Vision and Image Processing (CVIP) Ifeoma Nwogu Lecture 27 – Overview of probability concepts 1.
Static Equilibrium; Elasticity and Fracture
Chapter 12 Analyzing Semistructured Decision Support Systems Systems Analysis and Design Kendall and Kendall Fifth Edition.
1 Interpreting a Model in which the slopes are allowed to differ across groups Suppose Y is regressed on X1, Dummy1 (an indicator variable for group membership),
CSE Lecture 17 – Balanced trees
Preview Warm Up California Standards Lesson Presentation.
Lial/Hungerford/Holcomb/Mullins: Mathematics with Applications 11e Finite Mathematics with Applications 11e Copyright ©2015 Pearson Education, Inc. All.
Mani Srivastava UCLA - EE Department Room: 6731-H Boelter Hall Tel: WWW: Copyright 2003.
1/22 Worst and Best-Case Coverage in Sensor Networks Seapahn Meguerdichian, Farinaz Koushanfar, Miodrag Potkonjak, and Mani Srivastava IEEE TRANSACTIONS.
Multiple Regression and Model Building
January Structure of the book Section 1 (Ch 1 – 10) Basic concepts and techniques Section 2 (Ch 11 – 15): Inference for quantitative outcomes Section.
Area Of Shapes. 8cm 2cm 5cm 3cm A1 A2 16m 12m 10m 12cm 7cm.
9. Two Functions of Two Random Variables
Introduction Embedded Universal Tools and Online Features 2.
Copyright © Cengage Learning. All rights reserved.
Educational data mining overview & Introduction to Exploratory Data Analysis Ken Koedinger CMU Director of PSLC Professor of Human-Computer Interaction.
Learning from Learning Curves: Item Response Theory & Learning Factors Analysis Ken Koedinger Human-Computer Interaction Institute Carnegie Mellon University.
Notes on Logistic Regression STAT 4330/8330. Introduction Previously, you learned about odds ratios (OR’s). We now transition and begin discussion of.
Chapter 15: Model Building
1 Learning from Learning Curves: Item Response Theory & Learning Factors Analysis Ken Koedinger Human-Computer Interaction Institute Carnegie Mellon University.
Educational data mining overview & Introduction to Exploratory Data Analysis with DataShop Ken Koedinger CMU Director of PSLC Professor of Human-Computer.
Classification and Prediction: Regression Analysis
Warm Ups Preview 10-1 Perimeter 10-2 Circles and Circumference
Data Shop Introduction Ken Koedinger & Alida Skogsholm Human-Computer Interaction Institute Carnegie Mellon University.
Advanced Methods and Analysis for the Learning and Social Sciences PSY505 Spring term, 2012 February 6, 2012.
Area & Perimeter. Page ________ of Notebook Title: Area and Perimeter Title: Area and Perimeter Objective: Find Area and Perimeter Objective: Find Area.
Notes on Logistic Regression
Towards building a better cognitive model
Presentation transcript:

Learning Factors Analysis – A General Method for Cognitive Model Evaluation and Improvement Hao Cen, Kenneth Koedinger, Brian Junker Human-Computer Interaction Institute Carnegie Mellon University

Learning curve analysis by hand & eye … Steps in programming problems where the function (“method”) has two parameters (Corbett, Anderson, O’Brien, 1995)

Can learning curve analysis be automated? Learning curve analysis  Identify blips by hand & eye  Manually create a new model  Qualitative judgment Need to automatically:  Identify blips by system  Propose alternative cognitive models  Evaluate each model quantitatively

Overview A Geometry Cognitive Model and Log Data Learning Factors Analysis algorithm Experiments and Results

Domain of current study 15 skills 1. Circle-area 2. Circle-circumference 3. Circle-diameter 4. Circle-radius 5. Compose-by-addition 6. Compose-by-multiplication 7. Parallelogram-area 8. Parallelogram-side 9. Pentagon-area 10. Pentagon-side 11. Trapezoid-area 12. Trapezoid-base 13. Trapezoid-height 14. Triangle-area 15. Triangle-side Domain of study: the area unit of the geometry tutor Cognitive model:

Log Data -- Skills in the Base Model StudentStepSkillOpportunity Ap1s1Circle-area1 Ap2s1Circle-area2 Ap2s2Rectangle-area1 Ap2s3Compose-by-addition1 Ap3s1Circle-area3

Overview Cognitive Models & Cognitive Tutors Literature Reviews on Model Improvement A Geometry Cognitive Model and Log Data Learning Factors Analysis algorithm Experiments and Results

Learning Factors Analysis a set of factors that make a problem- solving step more difficult for a student Logistic regression, model scoring to fit statistical models to student log data A* search algorithm with “smart” operators for proposing new cognitive models based on the factors

The Statistical Model Probability of getting a step correct (p) is proportional to: - if student i performed this step = X i, add overall “smarts” of that student =  i - if skill j is needed for this step = Y j, add easiness of that skill =  j add product of number of opportunities to learn = T j & amount gained for each opportunity =  j p  Use logistic regression because response is discrete (correct or not) Probability (p) is transformed by “log odds” “stretched out” with “s curve” to not bump up against 0 or 1 (Related to “Item Response Theory”, behind standardized tests …)

Difficulty Factors Difficulty Factors -- a property of the problem that causes student difficulties  Like first vs. second parameter in LISP example above Four factors in this study  Embed: alone, embed  Backward: forward, backward  Repeat: initial, repeat  FigurePart: area, area-difference, area-combination, diameter, circumference, radius, side, segment, base, height, apothem Embed factor: Whether figure is embedded in another figure or by itself (alone) Example for skill Circle Area: Q: Given AB = 2, find circle area in the context of the problem goal to calculate the shaded area AB AB

Combinatorial Search Goal: Do model selection within the logistic regression model space Steps: 1. Start from an initial “node” in search graph 2. Iteratively create new child nodes by splitting a model using covariates or “factors” 3. Employ a heuristic (e.g. fit to learning curve) to rank each node 4. Expand from a new node in the heuristic order by going back to step 2

System: Best-first Search an informed graph search algorithm guided by a heuristic Heurisitcs – AIC, BIC Start from an existing model

System: Best-first Search an informed graph search algorithm guided by a heuristic Heurisitcs – AIC, BIC Start from an existing model

System: Best-first Search an informed graph search algorithm guided by a heuristic Heurisitcs – AIC, BIC Start from an existing model

System: Best-first Search an informed graph search algorithm guided by a heuristic Heurisitcs – AIC, BIC Start from an existing model

System: Best-first Search an informed graph search algorithm guided by a heuristic Heurisitcs – AIC, BIC Start from an existing model

System: Best-first Search an informed graph search algorithm guided by a heuristic Heurisitcs – AIC, BIC Start from an existing model

The Split Binary Split -- splits a skill a skill with a factor value, & a skill without the factor value. StudentStepSkillOpportunity A p1s1 Circle-area-alone1 A p2s1 Circlearea-embed1 Ap2s2Rectangle-area1 Ap2s3 Compose-by- addition1 Ap3s1Circle-area-alone2 StudentStepSkillOpportunityFactor- Embed Ap1s1Circle-area1alone Ap2s1Circle-area2embed Ap2s2Rectangle-area1 Ap2s3 Compose-by- addition1 Ap3s1Circle-area3alone After Splitting Circle-area by Embed

The Heuristics Good model captures sufficient variation in data but is not overly complicated  balance between model fit & complexity minimizing prediction risk (Wasserman 2005) AIC and BIC used as heuristics in the search  two estimators for prediction risk  balance between fit & parisimony  select models that fit well without being too complex  AIC = -2*log-likelihood + 2*number of parameters  BIC = -2*log-likelihood + number of parameters * number of observations

Overview Cognitive Models & Cognitive Tutors Literature Reviews on Model Improvement A Geometry Cognitive Model and Log Data Learning Factors Analysis algorithm Experiments and Results

Experiment 1 Q: How can we describe learning behavior in terms of an existing cognitive model? A: Fit logistic regression model in equation above (slide 27) & get coefficients

Experiment 1 Results: Skill Intercep tSlope Avg OpportuntiesInitial ProbabilityAvg ProbabilityFinal Probability Parallelogram- area Pentagon-area Student Intercep t student01.18 student10.82 student20.21 Model Statistics AIC3,950 BIC4,285 MAD0.083 Higher intercept of skill -> easier skill Higher slope of skill -> faster students learn it Higher intercept of student -> student initially knew more The AIC, BIC & MAD statistics provide alternative ways to evaluate models MAD = Mean Absolute Deviation

Experiment 2 Q: How can we improve a cognitive model? A: Run LFA on data including factors & search through model space

Experiment 2 – Results with BIC Splitting Compose-by-multiplication into two skills – CMarea and CMsegment, making a distinction of the geometric quantity being multiplied Model 1Model 2Model 3 Number of Splits:3 Number of Splits:2 1.Binary split compose- by-multiplication by figurepart segment 2.Binary split circle- radius by repeat repeat 3.Binary split compose- by-addition by backward backward 1.Binary split compose-by- multiplication by figurepart segment 2.Binary split circle-radius by repeat repeat 3.Binary split compose-by- addition by figurepart area- difference 1.Binary split compose-by- multiplication by figurepart segment 2.Binary split circle-radius by repeat repeat Number of Skills: 18 Number of Skills: 17 AIC: 3, BIC: 4, MAD: AIC: 3, BIC: 4, MAD: AIC: 3, BIC: 4, MAD: 0.075

Experiment 3 Q: Will some skills be better merged than if they are separate skills? Can LFA recover some elements of original model if we search from a merged model, given difficulty factors? A: Run LFA on the data of a merged model, and search through the model space

Experiment 3 – Merged Model Merge some skills in the original model to remove some distinctions, add as a difficulty factors to consider The merged model has 8 skills:  Circle-area, Circle-radius => Circle  Circle-circumference, Circle-diameter => Circle-CD  Parallelogram-area and Parallelogram-side => Parallelogram  Pentagon-area, Pentagon-side => Pentagon  Trapezoid-area, Trapezoid-base, Trapezoid-height => Trapezoid  Triangle -area, Triangle -side => Triangle  Compose-by-addition  Compose-by-multiplication Add difficulty factor “direction”: forward vs. backward

Experiment 3 – Results Model 1Model 2Model 3 Number of Splits: 4Number of Splits: 3Number of Splits: 4 Number of skills: 12Number of skills: 11Number of skills: 12 Circle *area Circle *radius*initial Circle *radius*repeat Compose-by-addition Compose-by-addition*area- difference Compose-by- multiplication*area-combination Compose-by- multiplication*segment All skills are the same as those in model 1 except that 1. Circle is split into Circle *backward*initial, Circle *backward*repeat, Circle*forward, 2. Compose-by-addition is not split All skills are the same as those in model 1 except that 1. Circle is split into Circle *backward*initial, Circle *backward*repeat, Circle *forward, 2. Compose-by-addition is split into Compose-by-addition and Compose-by-addition*segment AIC: 3,884.95AIC: 3, AIC: 3, BIC: 4, BIC: 4, BIC: 4, MAD: 0.075MAD: 0.079MAD: 0.077

Experiment 3 – Results Recovered three skills (Circle, Parallelogram, Triangle) => distinctions made in the original model are necessary Partially recovered two skills (Triangle, Trapezoid) => some original distinctions necessary, some are not Did not recover one skill (Circle-CD) => original distinction may not be necessary Recovered one skill (Pentagon) in a different way => Original distinction may not be as significant as distinction caused by another factor

Beyond Experiments 1-3 Q: Can we use LFA to improve tutor curriculum by identifying over-taught or under-taught rules?  Thus adjust their contribution to curriculum length without compromising student performance A: Combine results from experiments 1-3

Beyond Experiments Results Parallelogram-side is over taught.  high intercept (2.06), low slope (-.01).  initial success probability.94, average number of practices per student is 15 Trapezoid-height is under taught.  low intercept (-1.55), positive slope (.27).  final success probability is.69, far away from the level of mastery, the average number of practices per student is 4. Suggestions for curriculum improvement  Reducing the amount of practice for Parallelogram-side should save student time without compromising their performance.  More practice on Trapezoid-height is needed for students to reach mastery.

Beyond Experiments Results How about Compose-by-multiplication? InterceptslopeAvg Practice OpportuntiesInitial ProbabilityAvg ProbabilityFinal Probability CM With final probability.92 students seem to have mastered Compose-by-multiplication.

Beyond Experiments Results However, after split CMarea does well with final probability.96 But CMsegment has final probability only.60 and an average amount of practice less than 2 Suggestions for curriculum improvement: increase the amount of practice for CMsegment InterceptslopeAvg Practice Opportunties Initial Probability Avg Probability Final Probability CM CMarea CMsegment

Conclusions and Future Work Learning Factors Analysis combines statistics, human expertise, & combinatorial search to evaluate & improve a cognitive model System able to evaluate a model in seconds & search 100s of models in 4-5 hours  Model statistics are meaningful  Improved models are interpretable & suggest tutor improvement Planning to use LFA for datasets from other tutors to test potential for model & tutor improvement

Acknowledgements This research is sponsored by a National Science Foundation grant to the Pittsburgh Science of Learning Center. We thank Joseph Beck, Albert Colbert, and Ruth Wylie for their comments.

END

To do Reduce DFA-LFA.ppt, get from ERM lecture  Go over 2nd exercise on creating learning curves (from web site) in this talk & finish in 2nd session?  Print paper …. Other  Mail LOI feedback to Bett, add Kurt’s refs