|
Abstract:
|
Writing proficiency is essential to K-12 students; however, National Assessment of Educational Progress (NAEP) indicated that most U.S. students did not meet challenging grade-level expectation; furthermore, minority groups performed worse than mainstream students. This prompts the question: what are the characterizing variables and contextual variables for students who do not meet expectations? The study utilized a data-driven analytic approach—regression tree analyses—to display the patterns of the interplay among demographic factors and students’ writing experience with students’ outcomes among eighth- and twelfth-grade students who participated in 2011 NAEP writing. The subgroups as a result of regression tree analyses and the cluster structure (i.e., students nested in classes or schools) in the data served as two competing hierarchies for multilevel cross-classification modeling, which was studied to see whether it explained more variance in students’ writing performance than using multilevel modeling only. This study provides information on how subgroups perform so that resources can be provided, and pedagogies and interventions can be tailored to address students’ needs.
|
