Pre-Service Teachers’ Planned Accommodations for English Learners: An Assessment Approach for Systematic Improvement

As the dramatic increase in ELs in mainstream classrooms over the last decade continues to impact multiple facets of the educational system, teachers must respond and adapt to the complex and diverse needs of their students. To provide equitable instruction for ELs in the classroom teachers must understand the varied demands of EL instruction and plan appropriate accommodations to help these students grasp grade-level content. The present study investigated how intentionally pre-service teachers (PSTs) plan appropriate accommodations for English learners (ELs) to ensure comprehensibility of the content they are planning to teach. By examining lesson plans developed by PSTs using a rubric developed and approved by content experts this study provides insight into the level of preparedness with which PSTs finish their program. A Delphi Panel of experts completed multiple rounds of scoring written lesson plan accommodations anonymously using a rubric created for the project until a significant level (.90) of inter-rater reliability, or agreement amongst the three panelists was achieved. The classification data from the panelists was then used to train a supervised machine learning algorithm to score the written accommodations. Findings show that pre-service teachers primarily plan generic accommodations for ELs, starving them of the adequate instruction they deserve, and that machine learning can be used to help reduce the human workload for examining large, complex datasets in teacher education. This project serves as an important step in leveraging revolutionary technology for change in teacher education, and, more importantly, as a pathway forward in identifying appropriate instructional accommodations to reduce the inequitable instruction ELs receive.