Poster #70 - Academic Performance Among Elementary Students with Diverse Language Experiences

• In Event: 2-110 - Poster Session 07
  In Poster Session: PS 07 Section - Education, Schooling

Integrative Statement

Developmental research often compares monolingual and bilingual children in cognitive and language abilities. An open question is: how do these findings play out in education, such as on standardized state-level academic assessments? In the U.S., where English is the dominant language of instruction, unlike developmental research comparing bilinguals and monolinguals, comparisons are often made between those who are proficient in English and those who are English learners (ELs). A persistent achievement gap between these two groups is often reported (NCES, 2018). Simultaneously, we know that a significant portion of ELs are represented in low socioeconomic status (SES; Kieffer, 2010). Yet, we do not know how bilingualism and SES interacted and contributed to academic achievement. We conducted a secondary data analysis using state-level elementary education data from Massachusetts to investigate if (1) language experience is significantly associated with lunch status; and (2) language experience is associated with English Language Arts (ELA) and Mathematics state exam performance when lunch status is considered.

De-identified demographic data from grade 3 students from 2015-2016 was obtained from the state department of education (Total N = 57,281, excluding children in special education). Children were stratified by their language experiences: (1) native English speakers (NES); (2) English proficient bilinguals whose first language is not English and never received English learner services (EPBs); (3) Former English learners (FELs); and (4) Current English learners (CELs). We expected an achievement gap between CELs and NES. However, given adequate English proficiency, we expected the EPBs and FELs to have comparable academic performances relative to NES. The academic outcome variables were the ELA and Mathematics scaled scores as part of the Massachusetts Comprehensive Assessment System. These data were analyzed separately for those who were eligible for free-/reduced-lunch and those who were not.

Results showed a strong association between language groups and lunch status (χ2(3) = 6727.27, p < .001, see Figure 1). Moreover, children eligible for free-/reduced lunch had significantly lower performance in both subjects (Fs >1122, ps <.0001, see Figure 2). There was also a main effect of language group, Fs > 446, ps < .0001. Surprisingly, EPBs and FELs who were not low-income had comparable performance (ps > .09) and both groups had consistently better performance in ELA and mathematics than NES and CELs (ps < .020). Low-income FELs outperformed both EPBs and NESs in both subjects (ps < .001). Low-income EPBs and NESs performed similarly in ELA (p = .167) but EPBs outperformed NESs in math (p < .001). CELs, regardless of lunch status, had the lowest performance across academic subjects (ps < .001). Findings from this analysis call for increased collaboration between educators and researchers. To date, academic comparisons between English proficient and not proficient students have highlighted the shortcomings of English learners. However, the striking academic difference between English learners and their English proficient bilingual peers suggest that English learners also have the potential to be successful academically. Increased collaboration will inform research and practice, promote linguistic inclusiveness, and provide sensitive services for students from diverse language backgrounds.

Authors

• Gigi Luk, McGill University
  Presenting Author

• Brianna L. Yamasaki, Vanderbilt University
  Non-Presenting Author

• Jeannette Mancilla-Martinez, Vanderbilt University
  Non-Presenting Author