Project Work in the Computing Curriculum

1
Project Work in the Computing Curriculum

• Sally Fincher
• 15th Annual NACCQ conference
• Hamilton, New Zealand
• 3-5 July 2002

2
EPCoS

• Effective Projectwork in Computer Science (EPCoS)
• UK funded project (250k UKP)
• 3 years (1997-2000)
• 10 institutions
• Effective Project Work in Computer Science
  Principles and Pragmatics, Springer-Verlag 2001

3
Why projects?

• Learn something we cant teach but we want them
  to know
• Process
• extended period of time, pts, possible
  methodology prescribed
• Product
• often integrative of many aspects of previous
  curriculum (capstone), written and oral
  communication skills. What is a report, anyway?
• Industrial practice
• real projects, real clients, authentic
  experience, authentic assessment, motivation

4
Three features of project work

• Immersion
• Situation
• Location

5
Immersion

• Projects are free from normal curricular
  timetabling contact constraints (no lectures,
  self-directed)
• Students are free to do what they will with their
  time material (and often spend far too much
  time doing so)
• Immersion can be seen in individual projects just
  as much as team/group projects.

6
Immersion (i)
7
Immersion (ii)
8
Immersion (iii)

• My research library was in a shipping container
  rented out for self-storage. I got the steel
  8-by-8-by-40 foot space for 250 a month and
  spent all of 1000 fixing it up with white paint,
  cheap carpet, lights, an old couch, and raw
  plywood work surfaces and shelves. It was heaven.
  To go in there was to enter the
  book-in-progressall the notes, tapes, 5x8 cards,
  photos, negatives, magazine articles, 450 books
  and other research oddments laid out by chapters
  or filed carefully
• Stewart Brand How Buildings Learn, 1994

9
Immersion (iv)
10
Educational Examples

• Individual Research-type projects from
  final-year undergraduate to PhD
• Design-and-build projects (typically in groups
  and of a semester or more duration)

11
Three features of project work

• Immersion
• Situation
• Location

12
Situation

• Students have to learn differently when doing
  project work
• If situated learning (Lave Wenger) is an
  effective way to learn, can we use this in a more
  formal educational setting?
• The (frequent) interdisciplinary nature of
  project work combats the artificiality of the
  regular educational setting. Curricula divisions
  often have to be abandoned
• Project work is always contextualised

13
Educational example Aalborg

• Traditionally, higher education has been focussed
  on rule-based disciplines with independent
  identities in their own contexts
• Problem-oriented education, however, is based on
  working with unsolved, relevant and current
  problems from society/real life, e.g. the
  engineers' professional activity in an
  environment where solutions to real problems are
  sought.
• By analysing the problems in depth the students
  learn and use the disciplines and theories which
  are considered to be necessary to solve the
  problems posed, i.e. the problem defines the
  subjects and not the reverse.

14
Problem-oriented project work
The practical problem can be a symptom that
something is wrong with our theories and
assumptions, and thus the practical problem
produces a theoretical problem as to why there is
a practical problem
http//www.auc.dk/fak-tekn/aalborg/engelsk/problem
.html
15
Other educational examples

• PJ300
• Genesis at University of Sheffield

16
Three features of project work

• Immersion
• Situation
• Location

17
Location

• Classroom extension (catch-as-catch-can may
  have features of immersion, may not)
• Maximal colocation (also called radical
  colocation project rooms, war rooms)
• Distributed teams

18
Radical colocation

• One site collected productivity measures on the
  six teams that we observed, as they do with all
  their software engineering teams. The measures
  allowed a comparison of these groups with the
  company norm, which showed the company already
  well above (better than) the national average.
  The results were remarkable they produced double
  the function points per unit of staff time
  compared to the corporate average. They cut the
  total time to market (per function point) by two
  thirds, with none of the groups, again, even near
  the corporate average
• Gary Olson Judith Olson Distance Matters
• HCI 2000 vol 15 pp 139-178

19
Why does this work?

• could move from one sub-group to another
• overhearing gave awareness
• 42 flip charts over 6 weeks
• common referent (by gesture or glance)
• maintains critical information
• place for examining interactions and planning for
  new events
• record of chronology of ideas or associatively
  meaningful clusters

20
Educational example Roskilde

• Pedagogy was a crucial part of the experiment
  lectures, fixed syllabuses, examinations and
  strict admission criteria were swept away.
• There was to be consistent implementation of
  project pedagogy at Roskilde University students
  were to select topics that interested them and to
  carry out the work in groups.
• A project would be continuously assessed,
  examinations would not be necessary

21
Roskilde houses

• The project groups were organised in "houses",
  administrative and physical units that contained
  about 60 students, 5-6 teachers and a secretary
  during the first years. Working together, the
  teachers and students of the house decided on the
  themes for the project work. The intention was
  that the teachers' research, which was to deal
  with the same subjects as the students' project
  work, should take place in cooperation with the
  students.
• http//www.ruc.dk/ruc_en/about/RU-history/

22
Other examples

• Monash studio-based
• Very short-term projects one day, one week
• students need to be in the lab and part of the
  activity there
• Prof Ian Witten, NACCQ 2002

23
But is it authentic?

• What we expect of project work
• What we expect/assume of industrial practices
• Distributed teams - potentially globally
  distributed

24
Educational example Runestone

• Different specialities
• Different countries (and therefore different time
  zones)
• Different cultures
• http//www.docs.uu.se/docs/runestone/
• Have to deal with
• Technological challenges
• Common ground
• Context trust

25
Culture

• Mangers start with sweet talk the top of the
  hamburger bun. Then the criticism is slipped in
  the meat. Finally, some encouraging words the
  bottom bun. With the Germans, all one gets is the
  meat. With the Japanese, all one gets is the
  buns one has to smell the meat
• E.S. Browning, Wall Street Journal 3 May 1994
  Side by side Computer chip project brings rivals
  together, but the cultures clash

26
Educational example

• Brace yourself for problems. Do your work in
  time (an golden oldie, that one). Yes, when an
  American says I can do all this, it means in
  Swedish I will try my best (but no success
  guarantied)
• Mary Last, reporting student comment from the
  Runestone project, 1st CSERGI workshop, 1998

27
Back to more basic things

• Lots of ways to integrate project work.
  Immersion, Situation and Location are useful ways
  to think about whats going on, but not very
  useful when designing instances of project work
  in educational settings.

• Assessment
• Supervision
• Allocation
• Team/group projects
• Reflection
• Motivation

28
How to find things out how to use things

• In EPCoS we developed a projectwork map
• Great for us
• Pretty useless for anyone else
• Bundles
• Problem Statement
• This bundle is
• The way it works is
• It works better if (It doesnt work unless )
• Solution Statement

29
Red Card/Yellow Card (aka La Coupe du Monde
1998)

• Red card / yellow card (Give them a management
  tool)
• Students and staff alike are reluctant to reward
  group members who do not contribute. (Although
  some groups are perfectly happy to carry a
  hitch-hiker). In either case, it is impossible
  for staff to know precisely how much work each
  team member did only the students involved know
  this.

30
Bundle body

• This bundle gives students some control over the
  behaviour of members of their project group and
  allows their non-performance to be factored into
  assessment.
• The way it works is that students are allowed to
  issue others in their project group with yellow,
  and in extremis, red cards. A yellow card is
  shown to a student who is deficient in effort
  or attitude or in other ways not making a full
  contribution to the group and is then lodged with
  the project supervisor. Being shown a yellow
  card results in a known penalty being applied to
  the student (for example a fixed number of marks
  lost), though a yellow card may be cancelled by
  increased effort, or at a boundary between phases
  of the project, or after a set time. A student
  who attracts the maximum number of yellow cards
  can be shown a red card, which excludes the
  student from the rest of the project and sets the
  mark awarded to zero. There is no recovery from
  a red card.

31
Bundle conditionals

• It works better if staff set the parameters of
  control (the penalty, the number of yellow cards
  that can be carried)
• It doesnt work if the system leads to the
  frivolous use of penalties. It doesnt work
  unless day-to-day management of the resource/role
  allocation is in the hands of the group
  themselves.

32
Solution statement generalising the particular

• So find a mechanism which devolves some control
  over the performance of group members to the
  groups themselves.

33
More of the same

• http//www.cs.ukc.ac.uk/national/EPCOS