Cognitive Science and Education ACT R and the PUMP Tutor Praveen Paritosh CogSci 207 Fall 2004 Week 7 11 11 04 Material in this slide based on Ken Koedinger s invited talk at the 2001 Artificial Intelligence in Education Conference and Christian Lebiere s ACT R tutorial Education A great domain for Cognitive Science Model students reasoning To give the right feedback at just the right moment Create flexible easy to use environment Expressive rich multi modal interaction Intelligent tutoring systems are the textbooks of the future Pervasive as textbooks are today Not replacement for teachers textbooks More dynamic interactive effective model of learning than textbooks Influences teaching practice Support for lifelong learning Koedinger 2001 Cognitive tutors as modeling tool instructional model In Forbus Feltovich Eds Smart Machines in Education The Coming Revolution in Educational Technology Consequence ITSs will revolutionize education An ITS Success Case Cognitive Tutor Algebra aka Pump Most widely used ITS 50 000 students in 00 01 school year 200 000 projected 1700 Schools From the 2004 NYT article Exemplary Curricula by US Dept of Ed Most cited IJAIED paper in past 4 yrs Koedinger Anderson Hadley Mark 1997 Intelligent tutoring goes to school in the big city 300 Schools in 2000 01 Algebra Cognitive Tutor Cognitive Tutor Technology Use ACT R theory empirical studies of learners to create Cognitive Model Incorporates multiple strategies typical student misconceptions Strategy 1 IF the goal is to solve a bx c d THEN rewrite this as bx c d a Strategy 2 IF the goal is to solve a bx c d THEN rewrite this as abx ac d Misconception IF the goal is to solve a bx c d THEN rewrite this as abx c d Model Tracing Follows student through their individual approach a problem context sensitive instruction Knowledge Tracing Assesses student s knowledge growth individualized activity selection and pacing ACT R A unified theory of cognition Desiderata for a unified theory of cognition Philosophy Provide a unified understanding of the mind Psychology Account for experimental data Education Provide cognitive models for intelligent tutoring systems and other learning environments Human Computer Interaction Evaluate artifacts and help in their design Computer Generated Forces Provide cognitive agents to inhabit training environments and games Neuroscience Provide a framework for interpreting data from brain imaging Representational Assumptions of ACT R Declarative Knowledge Chunks Schema Frame like structure Has slots for various and an isa pointer for category membership Procedural Knowledge Production rules Conditions variables actions Main claims of ACT R 1 There are two long term memory stores declarative memory and procedural memory 2 The basic units in declarative memory are chunks 3 The basic units in procedural memory are production rules Declarative Procedural Distinction Declarative knowledge Includes factual knowledge that people can report or describe but can be non verbal Stores inputs of perception includes visual memory Is processed transformed by procedural knowledge Thus it can be used flexibly in multiple ways Procedural knowledge Is only manifest in people s behavior not open to inspection cannot be directly verbalized Is processed transformed by fixed processes of the cognitive architecture It is more specialized efficient Chunks Example 1 CHUNK TYPE NAME SLOT1 F SLOT2 SLOTN ACT3 4 isa ADDITION FACT ADDEND1 THREE ADDEND2 FOUR SUM SEVEN Chunks Example 2 Fact The cat sits on the mat proposition isa cat007 agent fact007 action sits on object mat A Production is 1 A 50 millisecond step of cognition 2 The source of the serial bottleneck in otherwise parallel system 3 A condition action data structure with variables 4 A formal specification of the flow of information from cortex to basal ganglia and back again Production Rules Describe How People Use Declarative Rules in their Thinking Declarative rule Production rules describe thinking patterns Side side side theorem Special condition to aid search IF the 3 corresponding sides of two IF two triangles share a side AND the other 2 corresponding sides are triangles are congruent THEN the triangles are congruent THEN the triangles are Using rule backward IF goal prove triangles AND 2 sets of corresponding sides are THEN subgoal prove 3rd set of sides Using rule heuristically IF two triangles look THEN try to prove any of the corresponding sides angles 100 Published Models in ACT R 1997 2002 III Problem Solving Decision Making I Perception Attention 1 Tower of Hanoi 1 Psychophysical Judgements 2 Choice Strategy Selection 2 Visual Search 3 Mathematical Problem Solving 3 Eye Movements 4 Spatial Reasoning 4 Psychological Refractory Period 5 Dynamic Systems 5 Task Switching 6 Use and Design of Artifacts 6 Subitizing 7 Game Playing 7 Stroop 8 Insight and Scientific 8 Driving Behavior Discovery 9 Situational Awareness 10 Graphical User Interfaces IV Language Processing 1 Parsing II Learning Memory 2 Analogy Metaphor 1 List Memory 3 Learning 2 Fan Effect 4 Sentence Memory 3 Implicit Learning 4 Skill Acquisition V Other 5 Cognitive Arithmetic 1 Cognitive Development 6 Category Learning 2 Individual Differences 7 Learning by Exploration 3 Emotion and Demonstration 4 Cognitive Workload 8 Updating Memory 5 Computer Generated Forces Prospective Memory 6 fMRI 9 Causal Learning 7 Communication Negotiation Group Decision Making Visit http act psy cmu edu papers ACT R Models htm link How Production Systems Fit into Cognitive Tutors The main step in developing a Cognitive Tutor is to develop a cognitive model Decompose the skill to be taught into small knowledge units We use production rules to represent these knowledge units A production system combines A set of if then production rules that transform data in working memory as directed by a procedure called the interpreter Cognitive Tutor Algebra Course Integrated tutor text and teacher training In computer lab 2 days week classroom 3 days week Learn by doing Project based Student centered Cooperative learning Teacher as facilitator Replicated Field Studies Controlled full year classroom experiments Replicated over 3 years in urban schools In Pittsburgh 60 Milwaukee 50 Results 50 100 better on problem solving representation use 15 25 better on standardized tests Koedinger Anderson Hadley Mark 1997 Intelligent tutoring goes to school in the big city Traditional Algebra Course Cognitive Tutor Algebra 40 30 20