Assignment And Quality Control Of Peer Reviewers
2001 Annual Conference Proceedings
Much work has been performed on assessing the validity and usefulness of peer assessment in the classroom. Much less effort has been invested in enumerating and classifying strategies for assigning reviewers, encouraging good feedback, and preventing clustering of grades. This study reviews the different approaches that have been taken to those problems. Several of these strategies are employed in PG, our Web-based application for peer review and peer grading. Usually, students are assigned
... ts are assigned randomly to review other students' work. Often, students work in teams, with each member of the team reviewing the other members. Or, students or teams may choose from a list of topics to work on. In this case, it is helpful to assign students to review others who have chosen the same topic. To encourage students to provide adequate feedback to their reviewees, several approaches can be taken. Students can be denied credit for the assignment unless they do the required reviews. Or, they can be prevented from seeing feedback on their work until they provide feedback to others. Multiple review periods may be employed, with students required to give some feedback in each period. A formula may be devised to allow reviewers to share in good grades received by their reviewees. Or students may be assigned to review each other's reviews. To improve the accuracy of grading, students can be required to pass a pre-certification test before being allowed to serve as peer graders. The instructor can supply a set of grading criteria, and discuss it with the students, either in advance or after the students complete their first round of review. Reviewer mappings can be constrained to assure that each student will review one paper from each quartile (etc.) of the class.