Impact of self-construal on rater severity in peer assessments of oral presentations

ABSTRACT Raters vary from each other in their severity and leniency in rating performance. This study examined the factors affecting rater severity in peer assessments of oral presentations in English as a Foreign Language (EFL), focusing on peer raters’ self-construal and presentation abilities. Japanese university students enrolled in EFL classes responded to self-construal questionnaires and gave and assessed in-class presentations. The results of path analysis revealed that peer raters’ self-construal and presentation abilities might be potential sources of subtle biases in peer assessments. Whereas peer raters who placed a high value on maintaining harmonious relationships with others were more likely to give lenient ratings over time, those who were sensitive about being evaluated by others tended to become gradually more severe in their ratings. Furthermore, a peer rater’s presentation ability positively predicted rater severity, indicating that individuals with better presentation abilities were more severe in their peer assessments.