Rationale
ENGR102 is a large course, typically involving about 600 students, which is currently taught using traditional means.
The students attend lectures in three streams. Tutorials are organized with one tutor communicating with a full classroom of
students. In tutorial, the students study solved problems, and are given homework consisting of similar problems to solve on their own.
In this kind of instructional setting, there is no opportunity for a student to get individual feedback from the tutors/lecturers.
 
Implementation (pedagogy)
Our goal is to improve students problem solving skills by providing students with an environment which will analyse
the student’s solution, maintain a model of his/her knowledge, and provide individualized feedback and support to the student.
Within ICTG, we have already implemented a number of constraintbased intelligent tutoring systems (ITSs), which have proven to increase students
learning significantly. Our experiences show that students are highly motivated when learning with ITSs, and achieve significantly higher results
in exams. For more information about our tutoring systems, please see the
ICTG home page.
 
Implementation (technology)
The system is a problemsolving environment. We assume that students
wil be familiar with the domain from lectures. The ITS offers many practice opportunities to
students. When solving problems, the student first draws the force diagram starting
from the problem text. The system checks the diagram, and provides feedback on any errors.
Once when the diagram is correct, the student can start computing the unknowns, by selecting an equation to use,
replacing the known into the formula, and calculating the value. The system evalautes the student's
solution and provides feedback on each step.
The ITS is being developed in ASPIRE, an authoring system that ICTG has developed. More information
about ASPIRE is available here.
 
Impact on Student Learning
The skill we are addressing is being able to solve problems in mechanics.
After interacting with the proposed system, the students will be able to:
 Draw the force diagram
 Select a goal for the next step
 Select a formula to use in order to compute a necessary unknown
 Substitue known values into a chosen formula
 Perform calculation
 Know when the problem has been completed
 On a higher level, relate problemsolving skills to declarative knowledge
The planned evaluation will focus on the following questions:
 Is the developed ITS effective?
 Does it support learning better than the traditional approach?
 Does the system increase students’ motivation?
We will compare performances of two groups of students.
One group will learn the material in the traditional way, via lectures and tutorials.
The other group will attend lectures, but the tutorials would be replaced by interaction with the system.
Time will be controlled. We will require students to sit pre and posttests, to measure their knowledge.
We will also collect data about their actions while solving problems, and analyze the data.
  
Quick Facts
Courses Impacted: ENGR102 Engineering Mechanics
Students Impacted: 600
Faculty Involved: Prof. Antonija (Tanja) Mitrovic, Dr Brent Martin, Dr Charles Fleischmann
Keywords: engineering
 
Contact Us
Tanja Mitrovic (tanja.mitrovic@canterbury.ac.nz)
Brent Martin (brent.martin@canterbury.ac.nz)
Charles Fleischmann (charles.fleischmann@canterbury.ac.nz)
 
Publications
 Mitrovic, A., Martin, B. & Suraweera, P. Intelligent Tutoris for all: the constraintbased approach.
IEEE Intelligent Systems, vol. 22, no. 4, 3845, July/August 2007.
 Mitrovic, A., Ohlsson, S. ConstraintBased Knowledge Representation for Individualized Instruction
Computer Science and Information Systems, vol 3(1), 122, June 2006.
(pdf)
 Mitrovic, A. and the ICTG team LargeScale Deployment of three intelligent webbased
database tutors. Journal of Computing and Information Technology, vol. 14, no. 4, 275281, 2006.
(pdf)
 Mitrovic, A., Suraweera, P., Martin, B., Zakharov, K., Milik, N., Holland, J. Authoring constraintbased tutors in ASPIRE.
M. Ikeda, K. Ashley, and T.W. Chan (Eds.): ITS 2006, LNCS 4053, pp. 4150.
(pdf)
 Baghaei, N., Mitrovic, A. A Constraintbased Collaborative Environment for Learning UML
Class Diagrams. M. Ikeda, K. Ashley, and T.W. Chan (Eds.): ITS 2006, LNCS 4053,
pp. 176186, 2006. (pdf)
 Suraweera, P., Mitrovic, A, Martin, B. The role of domain ontology
in knowledge acquisition for ITSs. In: J. Lester, R. M. Vicari
and Fabio Paraguacu (eds) Proc. 7th Int. Conf. Intelligent Tutoring
Systems ITS 2004, SpringerVerlag LNCS 3220, pp. 207216, 2004. (pdf)
 Suraweera, P., Mitrovic, A. An Intelligent Tutoring System for
Entity Relationship Modeling Int. J. Artificial Intelligence
in Education, vol. 14, no. 34, 2004, 375417.
 Martin, B., Mitrovic, A. Domain Modeling: Art or Science?
In: U. Hoppe, F. Verdejo & J. Kay (ed) Proc. 11th Int. Conference on
Artificial Intelligence in Education, pp. 183190, 2003.
 
