Computer Based Modeling for Engineering

1
Computer Based Modeling for Engineering
Dianne Raubenheimer, Jeff Joines, Amy Craig
Steve Roberts
2
Presentation Outline

• Project Background and Description
• Teaching Modeling and Problem Solving
• Highlights of Work with TE Department
• Assessment Overview Some Results
• Assessing Problem Solving
• Problem Solving Results

3
Background

• Arose out of work in the Student Owned Computing
  Program in College Of Engineering (COE)
• Objectives
• Examine how to develop students problem solving
  computational skills early in their program of
  study
• Link computational processes skills across
  courses in the curriculum (computational
  thinking thread)
• Increase student faculty engagement with
  technology

4
Computational Skills in the Curriculum

• Created new or revised existing courses in
  Textile and Industrial Systems Engineering
• Driven by need to develop students algorithmic
  thinking and problem solving using computers as a
  tool
• Reinforcing skills across the curriculum
• Developing departmental computational thinking
  thread

5
Sample Computational Thinking Thread Textile
Engineering
6
Teaching Modeling Problem SolvingAn Evolution

• Modeling and programming must be integrated into
  an introductory computing course (Discipline
  Specific)
• Microsoft Excel is an ideal vehicle to teaching
  modeling and programming (VBA) to build
  computer-based modeling/decision support systems
• Utilizing Tablet PCs and virtual office hours via
  IM
• Teaching with Student Owned Computers
• Immediacy Effect Students get immediate
  response on what works and more importantly what
  doesnt
• Using technology requires a paradigm shift in
  teaching in order for it not be a distraction

7
Teaching Modeling Problem SolvingIn-Class
Assignments

• Collection of in-class labs (Integrated Lecture
  and Lab)
• 27 In-class Labs (7 Pure Excel and 20 Integrate
  VBA and Excel to build decision support systems)
• Real-world case studies/problems presented in
  each lab
• Two to five different problems/ case studies
• Problems they will see later in other course
  (e.g., physics, chemistry, upper level
  engineering courses)
• Students given small tasks/steps leading them to
  solving the entire problem, answering questions
  along the way

8
Teaching Modeling Problem SolvingIn-Class
Assignments

• To be effective, in-class assignments must
• Occupy students during entire two hour class
  period
• Students are held accountable by having to answer
  a series of ?s to be graded each day
• Challenge students to think critically
• Allow students to seek help from teacher, TA
  and/or Neighbor
• Allow some flexibility for teaching moments
• Homework projects test problem-solving skills
• Real Data driven problems (Internet, Consulting,
  Research, Colleagues)
• Word problems with reflection questions

9
Typical Day in Modeling Class

• One PowerPoint Slide
• Overview of problems, topics of the day,
  announcements
• Students download todays spreadsheet and/or Data
• Begin the lab
• Each problem has little background followed by a
  series of steps
• Understand what can be done in a short amount of
  time (Snippets of Code)

10
Teaching Modeling Problem SolvingSummary

• Assessed the structure of new course at first
• Course content
• Effectiveness of the In-class labs
• Student Learning on real homework and projects
• Delivery mechanism utilizing Tablet PCs (Hybrid
  mode)
• Using Instant Messaging to Perform Virtual Office
  Hours
• After Reflection
• Students in Upper Level Courses Seemed Better
  Able To Analyze and Solve Problems
• Steps in the lab followed systematic approach to
  breaking down and solving problems
• Critical thinking on these problems, HWs, and
  projects

11
Assessment Overview

• We have assessment data relating to
• Confidence in using computational tools
• Views about learning with technology
• Views about introductory computer based modeling
  courses
• Attitudes and approaches to learning in
  introductory courses
• Factors impacting student success in these
  courses
• Focus for this presentation- The impact of
  technology on problem solving abilities

12
Confidence Using Technology Tools

• By the end of the introductory course, students
  significantly more confident in using
• 8/10 Excel tools and
• 10/10 VBA tools.
• It is important to reinforce concepts in
  subsequent courses
• confidence levels drop significantly by the end
  of the 300 level courses when the computational
  thread is not developed.

13
Views About Introductory Courses

• Survey data corroborates views expressed in
  open-ended questions
• Students value
• integrated lab and lecture (in-class labs)
• instructor demonstrating using computer
• small class size
• interaction with instructor, TA other students
• computer keeping them focused
• step-by-step nature of the activities
• practice and repetition (in-class homework)

14
Approaches to Learning in Introductory Courses

• Highest ranking statements (average all students)
  were
• Working through all the steps in homework
  assignments, and trying to understand the
  component parts is as important as the final
  solution, 
• I attend all class sessions so that I know first
  hand what has been covered in class, and
• Completing all the homework assignments will help
  me do well in this class.
• Lowest ranking statements included
• I like to allocate an amount of time to solving a
  problem and then stop after the time is up.

15
Approaches to Learning in Introductory Courses

• Final grades positively correlated with
• regularity of class attendance, and
• sticking with a solving a problem when it cannot
  be solved immediately
• Final grades negatively correlated with
• putting a problem aside when its solution is not
  immediately apparent, and
• only starting to work on homework a day or two
  before they are due.

16
Gaining more than computing

• Assessment of Individual course (TE/ISE 110)
• Course delivery mechanism improves student
  learning
• Students enjoy the course since a valuable set of
  tools is being learned
• Informal data shows impact on performance in
  internships/jobs
• Broader assessment purpose
• Conjecture that course and use of technology
  improves students problem solving ability
• Challenge How to assess the impact of the course
  and technology use on students problem solving
  abilities?

17
Problem Solving

• Used Wolcotts (2006) Steps for Better Thinking
  Framework
• Students completed an open-ended problem solving
  task outside of class and submitted all their
  work
• Students completed an online set of reflective
  questions relating to how they solved the problem
• Scored using Steps for Better Thinking rubric
  (next slide)
• Scored on task solution (slides 20-22)

18
(No Transcript)
19
Our Problem Solving Task

• Should you accept a new job or keep your current
  job?
• Current job provides 5K/month fixed income
• New job is based on 2 commission of monthly
  sales of equipment rentals 3K/mo fixed income
• Several other variables provided (ranges)
• Generic, open-ended scenario
• Students provide a recommendation then answer a
  set of questions to reflect on how they
  approached solved the problem

20
Scoring the Task Level 0,1

• Level 0 Cannot ascertain information, use
  equations correctly, or make a reasonable
  recommendation.
• Level 1 Calculate a point estimate and provide
  recommendation may recognize issues with their
  strategy but unable to account for them.

4,900 lt 5,000
21
Scoring the Task Level 2

• Level 2 Calculate best and worst case scenarios
  along with the most likely case. Provide a
  recommendation with a possible discussion of any
  risk and uncertainty associated with their
  decision.

4,900 6,600 4,220
22
Scoring the Task Level 3

• Level 3 Incorporate reliability by performing a
  simulation (examining various scenarios) to
  determine how often the new job will pay greater
  than 5K/mo.

73.6 to earn more
69.4 to earn more
23
Individual Score Sheet
24
Fall 2007 Problem Solving Task

• Generic problem solving task associated
  reflective questions implemented in 6 classes (5
  400-level 1 100-level)
• Score given for task problem solving
• Kept track of whether students used computational
  tools to solve problem (Students self selected
  whether or not to use technology)
• Results
• Students who used technology typically provided a
  more advanced solution (score 2 or 3/-)
• Students using technology had a higher level on
  the Wolcotts scale

25
Fall 2008 Activities

• Focused on problem solving
• Three 400-level sections solved the same problem
• Two sections were required to use computational
  tools
• Other section was left open for the students to
  decide
• Two 100-level sections solved same problem
• One section was required to use computational
  tools
• Other section was left open for the students to
  decide
• Followed same procedure for scoring tracking as
  in Fall 2007

26
ISE/TE 110 Fall 2008 Preliminary Results
27
ISE/TE 110 Fall 2008Preliminary Results (cont.)
28
400-Level Courses Fall 2008Preliminary Results
29
400-Level Courses Fall 2008Preliminary Results
(cont.)
30
Conclusions

• Students using technology were better problem
  solvers (Wolcotts rubric) and generated a better
  problem solution (task completion score)
• 100 level students performed at lower levels than
  400 level students - they do not have discipline
  specific knowledge
• Technology acts as an enabler (100 level)
• Understanding the problem better (taking it apart
  and modeling it)
• Technology acts as an enhancer (400 level)
• Makes it easier to do more analysis once the base
  case is setup (what if scenarios)

31
Q A
To view a complete presentation of this work go
to http//litre.ncsu.edu/VirtualPreCOE.html Emai
l cdrauben_at_ncsu.edu with questions