The Longitudinal Influence of Language of Testing on Reading Assessments for English Learners

Background: The demographic fabric of inner-city schools has drastically changed in the last decades. The Hispanic, Spanish speaking student population is the fastest-growing demographic of students across the country. Hispanic students represent the largest portion of student population in Texas, accounting for 52.4% of all students. Considerable empirical evidence shows that students identified as economically disadvantaged, Hispanic, and English Learners (ELs) struggle in reading and trail behind in reading performance throughout their schooling. Given this demographic shift and the continued struggles to meet standards in reading for many ELs, it is important to examine how language of assessment and language proficiency may influence reading performance. Purpose: The following three research questions were examined: (a) Is there a statistical difference in performance among third-grade bilingual students who took the State of Texas Assessments of Academic Readiness (STAAR) reading test in Spanish versus those who took it in English? (b) Is there a difference in STAAR reading performance in Grades 5 and 8 for students who took Grade 3 STAAR in English versus Spanish?; (c) How does language proficiency and reading performance in Grade 3 predict reading performance on high stake assessments (STAAR) in Grade 5 and Grade 8? Methods: The quantitative study utilized archival student data from an urban school district located in Houston, Texas. Data were collected for students identified as, Hispanic, Spanish-speaking ELs, served under a bilingual program in Title 1 schools who were, continuously enrolled from kindergarten to Grade 3. Outcome data were STAAR scores for Grades 3, 5, and 8 were included. A t-test was conducted to determine whether statistical differences existed between Grade 3 STAAR scores for students who completed the test in Spanish versus English. ANOVA was used to examine potential difference between these two groups in STAAR performance in Grade 5 and 8. Additionally, multiple regressions examined the influence of language proficiency and reading achievement in Grade 3, on STAAR reading performance in Grade 5 and Grade 8. Results: For RQ1, the t-test revealed a statistically significant difference (p<.01; ES=.23) between group on Grade 3 STAAR reading performance, in favor of the group who completed the assessment in English. The ANOVA for RQ2 showed that group testing in English continued to outperform the group testing Spanish in Grades 5 (p < .001; ES = .97) and 8 (p < .001; ES = .67). For RQ3, language proficiency evaluated by the Texas English Language Proficiency Assessment System (TELPAS) and reading performance STAAR in Grade 3 were strong and significant predictor of STAAR reading performance for both Grade 5 and 8. Conclusion: Students Grade 3 language and reading proficiency were key predictors of future reading achievement. The development of strong literacy and language skills in Grade 3 in both languages appears to translate to stronger reading achievement in upper elementary and middle grades. These findings support the connection that literacy and language development in the student first language (Spanish) impacts literacy and language development in the student’s second language (English). Further research is needed to evaluate ELs served in different language learning programs and analyzing the effectiveness of transitional bilingual programs. This research underscored the importance of reading programs in lower grades that enforce the literacy and language skills in both languages.