Evaluative misalignment of 10th-grade student and teacher criteria for essay quality: An automated textual analysis

Writing is a necessary skill for success in the classroom and the workplace; yet, many students are failing to develop sufficient skills in this area. One potential problem may stem from a misalignment between students’ and teachers’ criteria for quality writing. According to the evaluative misalignment hypothesis, students assess their own writing using a different set of criteria from their teachers. In this study, the authors utilize automated textual analyses to examine potential misalignments between students’ and teachers’ evaluation criteria for writing quality. Specifically, the computational tools Coh-Metrix and Linguistic Inquiry and Word Count (LIWC) are used to examine the relationship between linguistic features and student and teacher ratings of students’ prompt-based essays. The study included 126 students who wrote timed, SAT-style essays and assessed their own writing on a scale of 1-6. Teachers also evaluated the essays using the SAT rubric on a scale of 1-6. The results yielded empirical evidence for student-teacher misalignment and advanced our understanding of the nature of students’ misalignments. Specifically, teachers were attuned to the linguistic features of the essays at both surface and deep levels of text, whereas students’ ratings were related to fewer overall textual features and most closely associated with surface-level features.