Balancing Depth and Breadth in the Data Structures Course

1
Balancing Depth and Breadth in the Data
Structures Course

• Karina Assiter, Ph.D.
• Computer Science and Systems Department
• Wentworth Institute of Technology
• Boston, MA 02115
• 781.395.6922
• assiterk_at_wit.edu

2
Outline

• Motivation
• Data Structures
• In the CS Sequence
• Depth versus Breadth
• Techniques (Motivation, Implementation, Results)
• Beyond the classroom
• Pair programming
• Process-Log journals
• Web-based portfolios
• Complex projects
• Student-centered projects
• In the classroom
• Slides
• In-class exercises
• Group problem solving
• Results
• Conclusions
• Future Work

3
Motivation

• Data Structures course
• Spring 2003 (base case)
• Elaborate assignments with weekly deadlines
• Provide depth
• Analysis synthesis extrapolation
• Problem
• Many late submissions postpone deadlines
  limit breadth
• Pair programming
• Provide breadth
• Minimal late submissions cover full range of
  topics
• Problem
• Students rotate assignments limit depth
• What could we do?

4
Data Structures in CS Sequence
Data Structures (CS103)
Introductory Sequence (CS101/102)
Analysis of Algorithms
(Bridge)

• Purpose
• Introduce problem
• solving program
• design
• Selected Topics
• I/O (including file I/O)
• Selection and iteration control
• Logical operations
• Pointers, arrays, linked lists
• Functions
• Strings
• Recursion
• OOP concepts
• Dynamic Allocation
• Stacks and queues

• Purpose
• Introduce abstract data types
• (data structurestheir algorithms)
• Implement abstract data types
• ( rigorous programming).
• Selected Topics
• Stacks and Queues
• Linked list structures
• Trees
• Hash tables
• Indexed files
• Graphs

• Purpose
• Review basic algorithms
• Introduce analysis
• advanced algorithms
• Selected Topics
• Order of Growth
• Analysis notation
• Sorting algorithms
• Searching Algorithms
• Graph Algorithms
• Tree Algorithms

5
Challenge
Data Structures

• Depth
• Exposure to implementation issues, allowing
• Analysis
• Synthesis
• Extrapolation (to new problems)

• Breadth
• Exposure to range
• of data structures
• Trees
• Queues
• Graphs
• Stacks
• Lists

6

• Allocated vs Estimated
• Allocated Time
• Based on lecture
• Due to semester
• Estimated Time
• Based on projects
• Due to complexity
• For sufficient
• Analysis
• Synthesis
• Extrapolation

7
Choices

• Project time Coverage
• Less synthesis but with more exposure
• Project time Coverage
• More synthesis but with less exposure
• Project time Coverage
• Run over end of semester (not an option)
• Project time Coverage
• Preserve breadth without sacrificing depth?

8
Pedagogical techniques to address challenge

• Approach
• Complex projects
• Pair programming
• Process-Log journals
• Web-based portfolios
• Multiple Intelligence (Instruction)
• Student-centered projects

9
Preliminary Study

• Data Structures course
• Wentworth students
• Spring 2004
• Sample
• 20 students
• Analysis tools
• Qualitative
• Process logs
• Quantitative
• Questionnaire

10

• Approach
• Complex projects
• Pair programming
• Process-Log journals
• Web-based portfolios
• Multiple Intelligence (Instruction)
• Student-centered projects

11
Complex Projects

• Motivation
• Time-to-synthesize
• Complex problem solving
• Examples
• Comment Stripper
• Reverse Polish Notation (with stacks)
• Game of life
• Feasible with Pair-programming (next slide)
• programming resources
• dependence on external expert
  (instructor, TA)

12

• Approach
• Complex projects
• Pair programming
• Process-Log journals
• Web-based portfolios
• Multiple Intelligence (Instruction)
• Student-centered projects

13
Pair programming

• Motivation
• Pace projects with lectures
• Implementation
• Optional
• Encouraged, not required
• Informal
• No pre-defined roles, format
• Results (next few slides)
• Quantitative (questionnaire)
• Qualitative (process logs)

14
Pair Programming Questionnaire

• Self-selected to work with a partner
• Residential
• Extroverts
• Two brains better than one
• Assignments too difficult to do alone
• Agreed with without partner
• Assignments too difficult to do with a partner
• Agreed None
• Software Design Experience
• Working with partner made me spend more time
  planning how I would solve the problem
• Agreed with partner
• Neutral without partner
• Disagreed None

15
Pair ProgrammingQuestionnaire (continued)

• Emotional Support
• Working with partner increased my confidence
• Agreed 63
• Neutral 37
• Disagreed none
• Working with partner made me feel more confident
  about the material and the course
• Agreed 75
• Neutral 25
• Disagreed none
• Real-world Experience
• Working with partner was good practice for
  working in the real world
• Agreed All
• working with a partner improved my communication
  skills
• Disagreed none

16
Pair Programming Process Logs

• Two brains better than one
• Working with Peter, helps me to understand the
  assignments and the purpose.
• Software Design Experience
• Once I had them debugged my partner and I put
  our code together and started the full-blown
  testing phase
• Emotional Support
• Peter typed the syntax while I assisted in the
  thinking and problem solving part of the
  assignment.
• Real-world Experience
• Writing this code with a partner helped me
  improve my communication skills

17

• Approach
• Complex projects
• Pair programming
• Process-Log journals
• Web-based portfolios
• Multiple Intelligence (Instruction)
• Student-centered projects

18
Process Logs

• Motivation
• Monitor participation in pairs
• Additional benefits
• Monitor understanding of material
• Monitor communication and coordination
• Monitor cognitive development
• Analysis
• Qualitative Results (logs themselves)
• Quantitative Results (questionnaire)

19
Process logs Qualitative

• Understanding of material
• Doing this assignment was a nice review on
  templates, stacks, and queues for me.
• So far in my programming experience, I was
  taught about the concept of software re-usability
  but I never had to write a whole program that
  dealt with software re-usability (excluding
  header files). Thus, in a way, this assignment
  was a great experience for me.

20
Process logs Qualitative

• Communication and Coordination
• We started off by thinking about the thought
  process behind the problem before starting the
  code
• When we got to the lab, my partner and I first
  discussed how we wanted to write this program and
  who is writing what
• my partner and I split up the program so we
  would be doing an equal amount of work
• After testing the populated world with my
  partners print function, I had to figure out the
  best way to check whether each non-edge block
  should have an organism born, stay alive or die

21
Process Logs Quantitative

• Doing process logs helped me to think about how
  I solved the problems
• Agreed 60
• Neutral 40
• Disagreed None

22

• Approach
• Complex projects
• Pair programming
• Process-Log journals
• Web-based portfolios
• Multiple Intelligence (Instruction)
• Student-centered projects

23
Web Based Portfolios

• Motivation
• Accountability for programming products
• source code readability
• user interface usability
• executable files reliability
• Implementation
• Instructions
• promote your accomplishments by displaying on a
  web-based portfolio
• think of it as a programmers resume
• describe your work as if you are going to sell
  it

24
Web-based Portfolio Example
25
Web-based Portfolio Results

• Quantitative
• I had a web page before the class
• Yes 47
• No 32
• Sort of 21
• Portfolio improved quality of my assignments
• No indication
• Qualitative
• Students did not address the web-based portfolio
  requirement in their process logs.

26

• Approach
• Complex projects
• Pair programming
• Process-Log journals
• Web-based portfolios
• Multiple Intelligence (Instruction)
• Student-centered projects

27
Method of In-class Instruction

• Determines
• Learning rate
• Retention rate
• Translation into solutions
• Depends on
• Instructor
• Environment
• Interest
• Technology

28
Theory of Multiple-Intelligences

• We do not fully learn and retain information
  unless more than one of our intelligences is
  involved in the process.
• Becker, K., A multiple-intelligences
  approach to teaching number systems, Journal of
  Computing Sciences in Small Colleges, 19, (2),
  6-17, 2003.

29
Multiple Intelligence Instruction

• Context
• Data structures course at Wentworth
• Combined techniques multiple
  intelligences.
• Methods
• Slides
• In-class exercises
• Class developed solutions to programming problems
• Next slides Each technique with observed
• Benefits
• weaknesses
• conditions and co-techniques.
• Classification of learner types
• Becker, K., A multiple-intelligences approach to
  teaching number systems, Journal of Computing
  Sciences in Small Colleges, 19, (2), 6-17, 2003.

Accommodate
30
Method Slides

• Benefits
• Text linguistic learner.
• Illustrations and examples Visual
  learner.
• Handouts discussion focus (vs notes)
• Weaknesses
• Darkened room sleep
• Too many words hard to absorb
• Conditions and Co-Techniques
• Mediated classroom
• LCD display, computer, Powerpoint
• In-class exercises and class developed solutions
• actual development and implementation as a class
• Questions embedded into slides
• students must come to class for material, even
  when slides otherwise available to absent
  students.

31
Method Group/Pair Exercises

• Benefits
• Practice (learn by doing)
• Bodily-kinesthetic learner (physical activity)
• Interpersonal learner (social awareness and
  interaction)
• Weaknesses
• Variable completion rates
• un-productive conversation
• Less covered

32
Method Class Developed Solution

• Benefits
• Interpersonal learner (social awareness
  interaction)
• Learn from mistakes (ones own and others)
• Software Development Lifecycle
• Design, Implement, Compile, Debug, Test
• Weaknesses
• Student non-participation
• Time consuming
• Conditions and Co-Techniques
• Mediated classroom
• Safe environment
• Related to homework

33

• Approach
• Complex projects
• Pair programming
• Process-Log journals
• Web-based portfolios
• Multiple Intelligence (Instruction)
• Student-centered projects

34
Student-centered Projects

• Motivation
• Students internally-directed
• Qualitative Results
• The Game of Life .. will run each generation of
  the cells automatically, it is quite fun to
  watch! (students process log Game of
  Life assignment)
• RPN expressions interesting to solve
  (students process log reverse polish notation
  assignment)
• A nifty little program that simulates planes
  taking off and landing using queues!
  (students process log plane
  simulation assignment)
• Yet another enjoyable program to work on
  (students process log program comment stripper
  assignment)

35
Preliminary Study Selected Results (small
sample size, cant be unequivocally proven)

• Pair programming best with self-selection
• Worked alone
• indicated performed better alone
• Commuters
• did not select to work in pairs
• Worked in pairs
• Prepared for real world
• Improved
• communication
• confidence
• analysis/design skills

36
Preliminary Study Selected Results

• Slides with handouts
• Preferred over traditional board/notes method
• Process logs
• Monitor
• Pair programming
• Understanding
• Interest
• Encourage
• planning before coding
• In-class exercises
• Practice
• Limit absences

37
Conclusions

• Objective
• Results promising
• Results inconclusive
• Claims not proven due to small sample size
• Subjective
• Depth
• Complex problems assigned each week
• Assignments submitted on time
• Breadth
• All topics on syllabus presented before end of
  semester

38
Future Work

• Controlled Evaluation
• Two cohort groups
• Students using combined techniques, including
• Pair-programming
• Process logs
• Web-based portfolios
• Slides and in-class exercises
• Students not using the techniques
• Develop measures for quality of work between
• Web-based based group and control group
• Students working in pairs and students working
  alone

39
Future Work (continued)

• Evaluate grades at checkpoints
• Within and at end of semester
• Compare instructor contact hours
• Outside of class time
• Gather questionnaire results
• From each cohort group
• During the semester
• At later point
• How well did skills transfer to next course in
  sequence?

40
References

• 1 Amon, T., Bicycle club mileage log a servlet
  application for teaching web programming, Journal
  of Computing in Small Colleges, 19, (1), 218-225,
  2003.
•  
• 2 Anewalt, K., Beidler, J., Polack-Wahl, J.,
  Smarkusky, D., Group projects across the
  curriculum, Journal of Computing Sciences in
  Small Colleges, 19, (2), 232-237, 2003.
•  
• 3 Becker, K., A multiple-intelligences approach
  to teaching number systems, Journal of Computing
  Sciences in Small Colleges, 19, (2), 6-17, 2003.
•  
• 4 Chin, D., Prins, P., Tenenberg, J., The role
  of the data structures course in the computing
  curriculum, Journal of Computing in Small
  Colleges, 19, (2), 91-93, 2003.
• 5 Cliburn, D., Experiences with pair
  programming at a small college, Journal of
  Computing in Small Colleges, 1,9(1), 20-29, 2003.
•  
• 6 Clincy, V., Software development productivity
  and cycle time reduction, Journal of Computing
  Sciences in Small Colleges 19, (2), 278-287,
  2003.
•  

41
References (cont.)

• 7 Cockburn, W., Williams, L. The costs and
  benefits of pair programming, http//collaboration
  .csc.ncsu.edu/laurie/Papers/XPSsardinia.PDF.
  2004.
•  
• 8 DeClue, T., Pair programming and pair
  trading effects on learning and motivation on a
  CS2 Course, Journal of Computing in Small
  Colleges, 18, (5), 49-56, 2003.
• 9 Estell, J., Programming portfolios on the
  web an interactive approach, Journal of
  Computing in Small Colleges, 16, (1), 55-67,
  2000.
•  
• 10 Lewandowski, G., Using process journals to
  gain qualitative understanding of beginning
  programmers, Journal of Computing in Small
  Colleges, 19, (1), 298-309, 2003.

42
Questions?