Student 2018 PISA Reading Achievement: Identifying Predictors and Examining Model Generalizability for English Learners

• In Event: Poster 2

Abstract

Reading literacy is foundational for achievement across content areas, especially for English learners (ELs). However, reading research in the U.S. has focused mainly on English-only (EO) students in early/middle grades. We used machine learning to identify proximal and distal predictors of student PISA 2018 reading achievement. A total of 1482 variables (943, 347, and 192 at the student-, teacher-, and school-levels) served as the initial pool for variable selection. Among these, 21, 22, and 23 variables were selected using Elastic Net model as the most salient predictors of achievement. In the next steps—reported in the full paper—we will first estimate a hierarchical linear model on the entire sample and then test its generalizability to EL and EO populations.

Authors

• Tao Hao, Washington State University - Pullman

• Shenghai Dai, Washington State University - Pullman

• Yuliya Ardasheva, Washington State University - Tri-Cities

• Onur Ramazan, Chinese University of Hong Kong

• Robert William Danielson, Washington State University - Spokane

• Bruce W Austin, Washington State University - Pullman