Employing data from the first (2009) and second (2011) waves of the High School Longitudinal Study of 2009 (HSLS:09), this dissertation examines the relationship between mathematical literacy practices and U.S. Latina/os’ high school achievement. Additionally, this work investigates whether such relationship is influenced by students’ English-language background. HSLS:09 consists of nationally representative quantitative data from high school students as well as their families, teachers, classrooms and schools. Using an opportunity-to-learn conceptual framework, relationships between Latina/o students’ language, literacy and mathematics data are analyzed using a quasi-experimental design (QED) that incorporates principal components analysis, propensity score matching and multi-level linear regression techniques. This study proposes a quantitative operationalization of mathematical literacy practices and contributes insights about the relationship between content-area knowledge, instructional practices and language acquisition. Results inform policy and practice sequentially by providing evidence of the relationship between mathematical literacy practices in secondary mathematics and Latina/os’ mathematics achievement.

The relationship between a student's sense of belongingness and their academic achievement in high school has been well established in the research literature. However, few studies have explored this relationship among immigrant students. This study hypothesizes that belongingness and achievement for immigrant students is specifically mediated by peer and teacher influences. Therefore, engagement with academically inclined peers, and teachers who share an ethos of commitment toward their students, may result in both belongingness and achievement for immigrant students. In addition, a student's language status, which may help to develop social relationships, may also play an important role. Therefore, an investigation on the effect of peers, teachers and the English-Language status of immigrant students may clarify the link between belongingness and achievement. This dissertation uses the High School Longitudinal Study of 2009 (HSLS:09) to test a model that accounts for premigration measures including language status, while including the effect of peers, and teachers, on sense of belongingness and academic achievement. The analysis for this study is organized into two parts. In Study 1, belongingness is treated as an outcome and the effects of peers, teachers and English-language status are regressed on sense of belongingness. In Study 2, belongingness is treated as a predictor variable and the outcome is a student’s high school GPA. In the first study, the results showed statistically significant results for the effect of peers and teachers on sense of school belongingness. In the second study, the effect of peers, teachers and sense of school belongingness were statistically significant predictors of academic achievement. The limitations of this study and the implications for future research are also discussed.

Abstract

Teacher Profiles and High School Mathematics Achievement: What Do We Know About the Teachers of Latino and ELL High School Students?

Angela Thompson

Educational researchers have long sought to identify, measure, and explain the pathways to becoming a high-quality teacher, but specifying the precise admixture of preparation, knowledge, and policies remains elusive. Few studies link teacher qualities to student achievement, despite state and national calls for teacher accountability of their students' test scores (Guarino, Hamilton, Lockwood, & Rathbun 2006). Analyzing the characteristics of mathematics teachers whose students are achieving may help to diminish the long-standing opportunity gap in US schools between the large population of underserved students and those of the dominant culture and language. The persistent underperformance of Latino students, many of whom are also English Language Learners, is of particular concern.

The present study explored the qualities of teachers whose Latino and ELL students achieve in mathematics, and was guided by the following research questions: a) what are the characteristics of prevailing profiles of US 10^th grade mathematics teachers; b) what are the predominant characteristics of students who are assigned to 10^th grade mathematics teacher profiles; c) which teacher profiles, if any, are more likely to have Latinos or ELLs; and d) what combinations of matching students with teachers might predict better success for Latinos and ELLs in high school mathematics?

Using the ELS 2002–2004 longitudinal data set, a cluster analysis revealed five groups of teachers by their survey responses within a data set of over 4000 mathematics teachers. Results indicate that teachers whose students perform best in mathematics are primarily White, female, highly educated, regularly certified, and have many years of experience. Latinos and ELLs are significantly more likely to have newer, alternatively certified teachers who are Latinos themselves. For Latino and ELL students who scored well in mathematics, these newer, seemingly less prepared teachers may be best for them, but only if students are enrolled in high-tracked rigorous courses.

The results suggest more than one model of a high–quality teacher. Teachers should be highly educated in mathematics, but also highly prepared to teach special populations of students and to use the most recent technology in pedagogically appropriate ways. More importantly, students not enrolled in high-tracked, honors, AP, or college preparatory courses struggle in mathematics, regardless of their teachers' background or preparation.