MapleTA: A Smart Quiz System for Math, Science and Engineering
Paul Kates, Mathematics Faculty CTE Liaison
University of Waterloo, Nov 20, 2014
Overview
- what MapleTA can do now
- what MapleTA can do next year
What is MapleTA?
- a mathematically intelligent assessment system
- supports multiple question types
- algorithmic templates for questions
- grade management
- integrated with D2L (LEARN at UW, 2014)
Mathematically intelligent?
- many systems can't tell that 1/2 = 0.5
- MapleTA accepts numbers, expressions, formulas, functions, matrices, etc.
- power of Maple to create and grade questions (e.g., statistical
distributions, MathML rendering, plots)
- uses algorithmic variables
What are algorithmic variables?
- variables can be assigned random values
- set range of x from [-1,1] in 0.01 increments
- select a function from a list
- create a matrix with a certain rank
- generate a set of plots
- impose conditions
- x > y or (x+y > 0)
- rank != 0
- variables can be used anywhere in the question and the feedback
- diagram labels can also be algorithmic values
Algorithmic variable example
Assignment types
- Homework
- Anonymous
- Study Session and Mastery Assignments
- Proctored
Uses of MapleTA
- skill assessment, development, mastery
- pre-qualification
- pre-class warmup/check
- post-class review, post-module assessment
- independent self-study
- test, exam practice
- supervised tests, exams
MapleTA at UW
- free for UW students and courses
- math subjects: calculus, linear algebra, algebra, DE, finance, statistics
- science subjects: chemistry, physics(*)
- engineering subjects: circuits (ECE 140, GENE 140, MTE 120)
- engineering potential: math, material sci, thermodynamics, statics
- Math Readiness Test: Fall 2014: 1700 math students, 1700 engineering students
Content Development and Management
- you create and own the content
- modify shared content
- help, training and admin provided by Paul Kates (CTE), Sean Scott (MATH)
- questions are stored in central repositories, per subject
- courses inherit the relevant repository content
Lessons Learned
- within a repository, questions are organized by topic
- information fields include Author, Origin, Course, question Type, Difficulty,
and Keywords
- a tag on each question allows us to tell exactly where it came from when used
in an assignment
Lessons Learned cont'd
- preparing students in how to enter their answers is crucial
- guidance and training needed for instructors
- assignment policy consequences
- remarking assignments
- easy to access help
Student comments
- generally positive
- like availability of online systems in general
- wide range of login times and places
- motivated to keep trying for answer – persistance rewarded
- like immediate feedback – building confidence
- miss seeing intermediate steps in answer – a matter of partial credit
/