Dan:  Your experience reinforces the principle that faculty have the privilege of setting acceptable ranges for weighting IRAT/GRAT/application/peer evaluation BEFORE the students have opportunity to choose weighting.   For example, in our school of medicine, most faculty allow students to choose weighting within these ranges:

IRAT:  25-50%
GRAT: 30-60%
application:  0-30%
peer evaluation: 5-10%

I typically ask teams to decide among 4 options, each of which specifies a complete weighting scheme, such as "IRAT 30%, GRAT 40%, application 25%, peer evaluation 5%".  The teams make choices and defend those choices until there is a clear consensus--this sometimes requires a series of votes/discussions until one option is supported by a majority of the teams.

Dee Fink's alternative method is to use peer evaluation grades as multipliers of the team performance grades, so that peer evaluation matters, but is not assigned a specific percentage of the overall grade.  (for details, seen p. 267 in Team-Based Learning:  A Transformative Use of Small Groups in College Teaching; Michaelsen, Knight, and Fink eds, 2004, Stylus Publishing.

In any case, one definitely needs to limit the weighting assigned to peer evaluation so as not to diminish the importance of individual and team performance on the RAP and applications.   I would explain to the class that over-weighting of peer evaluation diminishes the affirmation of learning as the chief goal of TBL, so another grade weighting exercise is in order.     Good luck PK

    
Paul G. Koles, MD
Associate Professor, Pathology and Surgery
Wright State University Boonshoft SOM
937-775-2625 (phone)
937-775-2633 (fax)











On Jun 27, 2011, at 6:06 PM, Daniel Williams wrote:

Hi everybody:

I just went through the Grade-weight setting exercise outlined in appendix C of the TBL book with my class.  In previous semesters I had trouble getting classes of four teams to come to an agreement on grades, so for this semester's nine team class I used the large-class variant.  They set their weights individually and then entered them into an excel spreadsheet on my computer, where I had a running average for each category set up.  The problem is that the first team to finish entered in this:  10% individual performance, 10% team performance, 80% team maintenance.  I think these guys then persuaded the rest of the class to go along with them, so everybody else quickly gave me the same weights.  I was a little flabbergasted so I mentioned that this distribution was so crazy that a person could be really smart, but get dinged a letter grade for being overbearing or shy.  15 minutes later they had brought the team maintenance score down to 66%, but that still sounds really high to me.  Based on my experience the team and individual performance is usually split more or less 50/50 with team maintenance getting the remainder.

I tried to make the peer evaluation system simpler, no forced scoring, to minimize problems and I am worried that is what caused the stampede. 

Has anybody else run into this crazy result before?  I am a complete loss as to what to do about it!   

Thanks,
Dan Williams