Elementary English learners' engagement with automated feedback

Background: Despite the growing interest and use of Automated Writing Evaluation (AWE) in education, previous research has not examined how elementary-aged English Learners (ELs) in the United States interact with and benefit from feedback on their writing from these systems. Aims: The current study examines how elementary-aged students-categorized as proficient ELs, non-proficient ELs, and students ineligible for EL services (i.e., non-ELs)-engage with and benefit from the MI Write Automated Writing Evaluation (AWE) system in terms of behavioral, cognitive, and affective engagement. Sample: The study's sample consisted of 3459 students in Grades 3-5 from a school district in the Mid -Atlantic region of the United States. ELs comprised 23.7% of the sample (n = 820). Methods: After controlling for demographics and literacy achievement, the study applied regression and machine learning analyses to compare whether non-ELs, proficient ELs, and non-proficient ELs equally accessed MI Write's feedback, improved their writing, productively revised their writing, revised the same features of their writing, and positively perceived MI Write. Results: Language status was not related to students' behavioral and cognitive engagement with AWE feedback or their ability to benefit from that feedback. All students reported positive perceptions of MI Write, but proficient ELs held statistically significantly more positive perceptions than their peers. Conclusions: Study findings indicate that AWE systems like MI Write show promise for supporting the writing growth of elementary ELs.