| Learning outcomes of the assessment practice: | Questions to address:
1) Is peer assessment of assignments more harsh or more lenient than instructor assessment of the same assignments?
2) How do student-generated rubrics compare with instructor-generated rubrics?
Assessment and Class Details:
In an Earth Science common core course (CCST9013), poster presentations about improving Hong Kong’s living environment were evaluated by both students and instructors. Students were also asked to devise their own original rubrics to justify their peer evaluations. The class contained 100 students, and there were roughly 5 students per group for the making of the posters, rubrics, and assessments of other groups. Figures, instructor rubrics, and selected student rubrics can also be accessed at https://public.me.com/gavinwporter for further perusal.
Results:
1) Peer to peer grading resulted in a lower score in 14 out of 19 cases. The average instructor grade for the assignment was 76.7% (Standard deviation [SD] = 10.1) whereas the average peer grade for the assignment was only 68.4% (SD = 12.0).
2) Although some student-generated rubrics were quite detailed and impressive, the majority were much less specific than the instructor-generated rubrics. Out of 19 groups, only three of the groups had rubrics which were detailed and specific in regards to the assessment criteria.
Conclusions:
Students would most likely be made uncomfortable by harsh assessments with little justification, yet left to their own devices, they appear to be behaving in precisely this manner. This should lend some credence in student’s eyes to the care that instructors put into the assessment process, and should alert students that instructors are more lenient than their peers, by over a full grade level. This assessment resources website submission should be of interest not only to instructors, but also to students involved in peer assessment.
|
