Online

Morgan, Annabelle

College of Education and Social Services

2005

PhD

This dissertation project researched student mobility-- school changes not due to customary promotion-- and its educational correlates, for students and schools in Vermont. Student mobility research in other states has found that the majority of these students are disadvantaged youth from low-income families, and they lag behind their peers academically. Academic consequences of student mobility affect not only students, but also their schools since NCLBA implementation sanctions influence school enrollments by increasing student transfers. The need for information about rural student mobility during early NCLBA implementation is significant in predominantly rural Vermont. This was the first statewide study of outcomes of mobility for students and schools in a rural state. Three basic research questions were: (a) What is the incidence of mobility among Vermont students and schools? (b) What is the impact of mobility, i.e., how does the incidence of mobility vary according to educational correlates for students and schools? (c) What do analytical models that integrate student and school characteristics reveal about variation in mobility that may be useful for educational policy and practice? To address these questions, the study analyzed data for Vermont public school students, grades 1 through 12, during school years 1999-2004. Data sources included: (a) the Vermont Department of Education Student Census and Demographic Update; (b) student New Standards Reference Examination English Language Arts and Mathematics tests, grades 4,8, and 10; (c) Vermont School Report indicators, and (d) NCES-US Census public school location information. In-depth cross-sectional and longitudinal analyses of mobility, performance, sociodemographic, and educational correlates revealed significant and disturbing relationships that merit policy and prevention follow-up programming. School-level mobility incidence indicated that while in-migration was 20% on average, over 30% of the schools experienced much higher rates, mirroring urban-based mobility incidence. Academically, mobile students performed 3-10 percentile ranks lower than their stable counterparts did across grade levels and content areas on standardized tests, for longitudinal cohorts as well as cross-sectional grade groups. Risk factor analyses revealed that mobile students, relative to their stable peers, were (a) more likely to participate in free or reduced lunch programs at school, (b) less likely to have a 504 plan in place, (c) more likely to have kept a writing portfolio for 0-1 years (versus 2-5 years), (d) more likely to have kept a mathematics portfolio 0-1 years (versus 2-5 years), and (e) more likely to not meet the standard on mathematics performance tests. Hierarchical generalized linear modeling analyses indicated that between 8% and 32% of the variation in student mobility was attributable to school-level composition and resources. This project aimed to benefit the Vermont educational community in several ways. Analytical methodology will provide the framework for developing a longitudinal monitoring system with mobility incidence, impact, and relevant educational information. Information from analytical results will inform a case study during spring 2005 to address student mobility by raising public awareness of associated issues that affect not only the students and their families, but also classrooms, schools and communities.