Investigating the reliability and usefulness of self- and peer assessments of a capstone design project

Abstract

There is increasing use of self- and peer assessments to assess behaviours of students working on group projects. This study aimed to explore the reliability and usefulness of self- and peer assessments during a capstone design project. A sample of 61 final-year undergraduate students aged 23 to 25 years old who were enrolled in Bachelor Degree of Chemical Process Engineering participated in the study. Students worked in groups of 5 to 6 members for 28 weeks to complete the project. Training was provided, and progress was monitored. Self- and peer assessments were conducted during the 6th, 14th and 22nd weeks. In each assessment, students rated their own behaviours and those of their peers using identical Likert scale questionnaires, and they also wrote feedback to themselves and their peers. Quantitative findings reported that, in the 6th week, students ranked themselves (mean = 3.98) significantly lower than how they ranked their peers (mean = 4.16). In the 14th week, students still ranked themselves (mean = 4.14) lower than how they ranked their peers (mean = 4.20). Last, in the 22nd week, students ranked themselves (mean = 4.24) equivalent to how they ranked their peers (mean = 4.24). For qualitative findings, feedback written to peers in the 22nd weeks was compared to self-assessment feedback from that week. Self- and peer observations on one’s strengths and areas for improvement seem to converge both quantitatively and qualitatively towards the end of the project. It is also noted that both self- and peer assessment scores increased between the first and third assessments. The findings imply that students’ behaviours improved while working on the capstone project. In conclusion, self- and peer assessments could be reliable and useful for chemical engineering students, and training students in how to conduct these assessments is essential to ensure successful implementation. Future qualitative research could identify how and why students gradually change their behaviours in long-term, team-based projects.