CPSC 481: Evaluation Of Group Work

Description

Based on my past experiences from teaching this class, most students have found group work to be a positive experience.  Not only does having additional people working on the project reduce the pressure (and workload), many students also find that it produces a better project: in terms of design having several people working on the project helps spur the creative process as multiple people are able to come up with a more diverse set of ideas for prototype interfaces; in terms of evaluating the prototype, different people tend to spot different weaknesses in the design so again a better end product is produced.  

Unfortunately for some groups not all group members contribute an equal share of the workload.   For students who have taken the course in the past, I have gotten requests from the people who have had to "carry" their group members for me to develop a group evaluation mechanism to help prevent this from happening to students in the future.  There are two parts to the group evaluation mechanism that I came up with:

1. A contribution form.
2. Numeric group evaluations.
   

1) Contribution forms:

The contribution forms indicate who did what for each assignment.  To complete this form you must sit down each week as a group and determine which group members contributed to what parts of the project.  This process is not an evaluation process but merely stating the facts of the situation so there is not need for anonymity.  You can download the forms, which you find for each assignment component and simply fill in the applicable boxes for each person.  If a person worked on a particular assignment component then put a check mark in the box.  It is important that this process be done as a group - make sure that all group members are involved!   All group members must view and sign each page of each form or your TA's will not provide you with a mark for your project.

e.g., Suppose that we have a group of three students: Larry Smith, Moe Smith and Curly Smith.  These three students write up a list of 5 task examples for their project portfolio:

• The first task was written up by Larry alone.

• The second task was written up by Moe alone.

• The third task was written up by Larry alone.

• The fourth task was written up by Larry and Moe together.

• All the group members contributed equally to the writing up of the fifth task.

Every time that a group member was involved a substantial portion of the work, that person receives credit for having worked on that portion of the assignment.

Exert of the contribution form for Larry, Moe and Curly

By itself the contribution form won't have an effect on the grade.  I won't have the time or inclination to interfere with each group's work processes unless I am flagged down that there is a problem which you can indicate to me in Part 2: The numeric group evaluations.  Put your completed contribution forms at the very front of your binder when you hand in your assignment.  For students who are worried that their group members may offer to "help" with a section of the project but who do no work I suggest that you only have one person work on a portion of the project.  For example with Larry, Moe and Curly you avoid having more than one person work on a task say by requiring that each group member come up with 2 or 3 task examples. 

If you group members all find that everyone consistently puts in the same amount of work each week so you don't feel that there is a need for indicating time and time again that things were equal then you can ignore the tables on the first few pages of the contribution forms and simply skip to the last page:

Approach #2: Short Form (use this as alternative if your group feels that the contributions were roughly equal)

In this case all your group has to do is to list the names of all group members and have each person sign below their names.  You can also mix and match between the two from week to week depending upon your situation e.g., for all weeks except the first everyone contributed equally so you would use approach #2 for the second week onwards and fill in the full table of values for the first week.  If you take the second approach that's great!, ideally I really like semesters when all groups of all groups are putting in an equal amount of work because it shows an interest in the class and (hopefully) it means that students will tend to do better that term.  Make sure that this if your group says that the contributions were equal that they really were.  If all I see in your portfolio is five weeks of equal contributions then it won't be possible for me to make any adjustments even if you send me email (see point #2 below) indicating otherwise because there won't be any evidence to support your claims.

2) Numeric group evaluations

These evaluations are to be emailed to the course instructor and to the tutorial TA whose tutorial that your group is attending.  Make sure that the subject line of your email is titled as follows: 

• Assignment 1 Subject line: "Assignment 1 numeric group evaluations"

• Assignment 2 Subject line: "Assignment 2 numeric group evaluations"

• Assignment 3 Subject line: "Assignment 3 numeric group evaluations"

The numeric group evaluations is an anonymous evaluation to be completed by each group member of all the group members.  (You must include yourself in the evaluation).  The required information includes the names of each group member and an integer value indicating how many letter "steps" that a particular person's grade should be adjusted by (e.g., "A" to "A-" is a letter step).  Restrictions on the numeric steps are:

• This process is used to indicate the proportion of work that a particular person contributed to each assignment relative to the contributions of the others.   So if one person contributed a smaller portion that means that another person contributed a larger proportion of the total work put into the project.  Consequently the sum of all grade adjustments must be zero - otherwise group members could simply give each the maximum grade and evaluation process would be meaningless.
• The highest possible grade is an A+ (4.3) regardless of how many letter steps that the group members want someone's grade to be increased.  
• The lowest possible grade is an F (0.0)  regardless of how many letter steps that the group members want someone's grade to be decreased.
• Adjustments must be an integer value - no fractional values are allowed.  If the difference in the amount of work contributed is too small to be measured by a grade step then it probably isn't worth bothering with.

Example from above