Peer-to-Peer Assessment in Large Classes: A Study of Several Techniques Used in Design Courses

Peer-to-peer assessment, with student teams evaluating the work of other teams, has the potential to greatly enrich student learning and the classroom environment, as well as compliment instructor-based assessments. This paper examines four peer-to-peer assessment techniques used in large engineering design courses. These techniques have been identified, and, through several iterations, developed to work well in the design engineering classroom setting. The four techniques include: evaluation of designs through a Poster Gallery Walk; Design Tournaments where candidate designs are successively eliminated through multiple rounds of review; “Hot Seat” Debates based on a ranking of student designs along a common performance metric; and “Design Manager Role Playing” where teams authorize or reject designs created by their peers. Based on survey responses, student impressions of the activities are mixed, but more students than not report that they feel the activities are effective in improving understanding of the course content and in developing critical analysis and evaluation skills. Activities in which students were given time to examine a small number of peer designs in depth, as opposed to a quick assessment of a large number of designs, were more positively regarded. The student assessments were generally found to correlate well with instructor and teaching assistant assessments (up to r = 0.67), similar to as has been reported elsewhere in the literature. By carefully phrasing the peer-to-peer assessment activities, and then evaluating the quality of the student assessments, it is believed the activities can be used to demonstrate, in part, several desired attributes for outcomes-based accreditation (such as problem analysis, investigation, design, and communication, for example).