University wide course evaluation pilot project

1
University wide course evaluation pilot project

• Report to Senate
• Tuesday, March 07, 2006

2
Sample
3
Means for Student Survey
4
Summary of Student Comments

• Background Questions concerns about maintaining
  anonymity
• Rating Scale
• Wanted middle value (neutralizing effect)
• Objected to negatively worded items
• Behavioral Scale
• Commented both positively and negatively on
  format
• General written comments are most important
  dont use class/computer

5
Faculty Feedback
6
Faculty Comments

• I see pros and cons to each
• Comments on individual items on both
• Emphasized space for comments
• Negative questions on rating scale a problem
• Output feedback pros and cons

7
Comparison of Results

• Two largest classes sampled
• In class with overall pattern of positive
  responses, consistency in the performance of the
  two forms
• In class where rating items showed bimodal
  pattern, behavioral items were similarly
  distributed

8
Sample Behavioral Scale Output
9
Sample Rating Scale Output
10
Differences

• Behavioral Scale
• Provides list of descriptors common language
• Shows distribution of responses
• Elicited many more comments from students

• Rating Scale
• Provides number
• Bimodal distributions are misrepresented in
  summary statistics
• Elicited fewer comments

11
Recommendations

• Option A
• Behavioral core with depart/course specific
  additions
• Report full descriptors
• Core used periodically
• Whatever is chosen, important to educate about
  how information is analyzed and used

• Option B
• Rating core with depart/course specific additions
• Report percentages of all responses, not just
  mean
• Use uniform scale sa-sd (1-4 or 4-1) for scaling

12
Thanks to .

• Course Evaluation Subcommittee for long hours and
  thoughtful discussions
• Participating faculty and students for their time
  and thoughtfulness
• Jeanne Wielgus for consultation and data analysis
• Rick Vrem, Kristine Nelson and Jodi Baker for
  assistance and support