A quasi-experimental study of formative peer assessment in an EFL writing classroom

Abstract

The benefits of using formative peer assessment have been broadly recognized by educators. However, in the context of English as a Foreign Language (EFL) teaching, research into the effectiveness of peer assessment has produced mixed results. One major focus of research is how training can influence the effectiveness of peer assessment. A number of case studies have reported positive outcomes using particular training methods, but there is relatively little research into the application of such training methods and their effects in ordinary EFL classes. This thesis reports a study of training in peer assessment in the context of a typical EFL class. The details of the training are explored in some depth, including evidence of its effects on the nature and quality of peer assessment and the reasons for those effects. A quasi-experiment was designed to investigate the effectiveness of the extensively discussed ‘Step Training’ proposed by Min. Data were collected in the form of students’ written feedback, essays, video recording, questionnaires and interviews. The nature of students’ feedback and the quality of their essays were explored. Classroom interactions were analysed and interpreted using socio-cultural theory. Analysis of the questionnaires revealed students’ attitude towards the training and perception of their capabilities in working on peer assessment. Student interview data was analysed thematically. The findings provide a complex picture of the peer assessment training. First, the ‘Step training’ appeared to lead to improvements in the quality of both essays and peer feedback. Secondly, analysis of the classroom discourse suggested that socio-cultural factors had both positive and negative effects on the students’ learning. Moreover, ANCOVA analysis of psychological features such as attitudes towards and perceptions of the peer assessment training suggested that the students were less influenced psychologically by the training. Finally, analysis of interviews identified students’ concerns about the design of the training course. In conclusion, it is suggested that for a typical EFL writing class, the effectiveness of a rigid training method such as ‘Step training’ should be reconsidered to take into account the influence of socio-cultural factors in classroom interaction. Rather than relying just on students’ end products to monitor the effectiveness of training, socio-cultural interaction should also be examined, as this is important in developing an identity as an effective peer assessor. Suggestions for improving the design of peer assessment training are provided.