Universal screening with automated essay scoring: Evaluating classification accuracy in grades 3 and 4
Wilson, J. (2018). Universal screening with automated essay scoring: Evaluating classification accuracy in grades 3 and 4. Journal of School Psychology, 68, 19-37. https://doi.org/10.1016/j.jsp.2017.12.005
This study examined whether a direct assessment of writing scored with automated essay scoring could be used as a universal screener to identify third- and fourth-grade students at risk of failing the Smarter Balanced English language arts assessment. Using ROC curve analysis and logistic regression, the study evaluated classification accuracy for fall and spring screening data.
Key takeaways
- Automated writing screeners showed acceptable to excellent classification accuracy, with stronger results in grade 4 than grade 3.
- Sensitivity-based and d-based cutpoints were reported to support screening decisions.
- The findings suggest that automatically scored direct writing assessments may be useful for early identification of writing-related academic risk.
Why it matters
Provides validity evidence for using automated essay scoring within universal screening and response-to-intervention frameworks for elementary writing assessment.
Coauthors
Topic tags
External links