Path: utzoo!utgpu!news-server.csri.toronto.edu!qucis!dalamb
From: dalamb@qucis.queensu.CA (David Lamb)
Newsgroups: comp.edu
Subject: Re: Group projects
Message-ID: <806@qusunitg.queensu.CA>
Date: 20 Jul 90 16:56:17 GMT
References: <1990Jul17.120036.8944@pdn.paradyne.com> <804@qusunitg.queensu.CA> <1990Jul19.215334.28746@nmt.edu>
Organization: Queen's University, Kingston
Lines: 44

In article <1990Jul19.215334.28746@nmt.edu> john@nmt.edu (John Shipman) writes:
>The toughest situation is when a team failed because of one
>member, yet the others worked hard and did good work.
>Clearly, the offender should get a low grade.  But should we
>also penalize the people that worked hard?  I can think of
>arguments both for and against this.
In a school environment, where everyone seems so concerned about grades these
days, you ought to do something to avoid hurting the people who worked hard.
I have handled this in the past by
1. Having teams of 4-5 people and setting the workload to what I think
   3-4 can do.
2. Have the students plan a series of subsets.  This gives each team a
   fallback position.
3. Having a series of deliverables, such as user's guide and preliminary
   design.  The team is less likely to fail, since you can detect poor
   performers earlier.
4. Once someone copped out on the last major deliverable (final running
   system), and the rest of the team found itself unable to deliver any
   working system.  I examined their code far more carefully than I usually
   do, to see how far they were from working, so I could give them some sort
   of partial grade.  I then graded the dropout as though he'd been a
   1-person team that handed in the real team's deliverables, except getting
   a 0 on the last one.  I then raised the real team's grade by a fraction
   that corresponded to the difference between the dropout's grade and the
   initial partial grade.
I can't really justify point 4 except on fairly vague grounds, but the
students all accepted it as reasonable.  Taking into account the
individual work, which wasn't hurt by the dropout, the team got roughly
B+ grades overall when others were getting B+ or A-, so didn't get hurt
too badly.  The dropout got a C; he had contributed to all those other
deliverables, so I didn't think I should fail him.

All this presumes you're talking all-term team projects.  If you're
doing a whole bunch of different team assignments that don't build on
each other in series, it should be possible to assign people to
different teams for each assignment, and assign individual grades based
on analysis of variance on the team grades.  I haven't tried this;  I'd
be interested in comments.

David Alex Lamb			ARPA Internet:	David.Lamb@cs.cmu.edu
Department of Computing				dalamb@qucis.queensu.ca
    and Information Science	uucp:   	...!utzoo!utcsri!qucis!dalamb
Queen's University		phone:		(613) 545-6067
Kingston, Ontario, Canada K7L 3N6