The design and application of a web-based self- and peer-assessment system

This article describes the motivation for and design of a web based self- and peer-assessment system, and reports on a study with 76 students (14–15 years old) in a course on web design which used the system.

The motivation for the development of the system was to better support proposing and reflecting on criteria for assessment, as well as supporting self- and peer-assessment. As previous research has demonstrated, these processes have beneficial effects on cognitive and meta-cognitive as well as affective aspects of students’ learning. The authors argue that the web based system offers several advantages over a more conventional paper-based implementation of peer-feedback exercises. The web based system’s document handling is more convenient then shuffling papers, it is more efficient in compiling scores and comments, and it provides more opportunities for peer-interaction in large classes. A teacher can quite effortlessly adapt an assessment exercise based on the web based system to his or her liking, e. g. set group sizes or configure the different phases of the exercise.

In their classroom study, the authors focused on what they call progressively focused self- and peer-assessment. This scenario unfolded in the following phases: (1) Each student submitted a web page. (2) Each student self-assessed his or her own work. (3) Students are randomly assigned to small groups and peer-assessed the works of the other group members. After the peer-assessment, they could modify their self-assessments. (4) A sample of “best” and “poorest” ranked works from other groups was randomly assigned to each student for assessment. (5) The overall “best” and “poorest” ranked works were presented to all students. (6) The students were given the opportunity for a last revision of their self-assessments. (7) The teacher reflected on the process in a discussion with the class.

For the purposes of evaluating the process, two experts (other teachers not involved in the process) rated all students’ works twice, once after the students first submitted their works, and again the students submitted their revised works. The study found that there is a tendency for the discrepancy between students’ self-assessments and experts’ assessments to decrease over the phases. Based on the experts ratings, the study also found that the quality of students’ works significantly improved. Last but not least, the study confirmed what other studies have shown before, i. e., that students’ assessments and experts’ assessment are remarkably consistent.

The authors describe a very interesting tool and a self- and peer-assessment process which progressively helped students better evaluate their own work. Systems of this type might well be of interest to many teachers. It is a pity that the prototype developed for and used in this study is not available on the web for others to use.