A sample of Santa Ana Unified School District (SAUSD) 8th grade students (N = 4,320) was selected to examine the pervasiveness of student mobility and its impact on academic achievement and English language development. Using descriptive statistics and multivariate analysis including analysis of variance (ANOVA), multiple regression and discriminant function, this study sought to examine mobility within an urban school district with a large Latino and English learner (EL) population.
Descriptive statistics revealed that at the district level, 70.7% of students were "stabile," having remained at their school during their intermediate school experience, 21% of students moved once and 8.5% of students were "highly mobile," having moved more than two times over three years. At the aggregate level, students appeared to be relatively stabile. But, at the disaggregated level, it became apparent that stability varied considerably depending on the program and the site. Fundamental schools had between 3.2% and 5.9% mobile students, whereas non-fundamental schools had percentages exceeding 32.6%. This finding points to the importance of viewing mobility data both at the aggregate and the disaggregated level, as this analysis provides a more accurate picture of mobility within a district.
Descriptive statistics also revealed differences in standardized test results and in the distribution of males and English learners among stabile, mobile and highly mobile student groups. According to the results, mobile and highly mobile students had lower over-all performance on the California Standards Test (CST) in English language arts (ELA) and mathematics than their stabile peers. Within-group analysis also revealed that a disproportionate percentage of male students were represented in the mobile and highly mobile student groups. Within-group analysis of English learners indicated that there was an uneven distribution of English learners among stabile, mobile and highly mobile student groups. The more students moved, the more likely they would be English learners: 34.6% of stabile students, 58.5% of mobile students and 69.1% of highly mobile students were English learners. This information provides evidence that differences exist among students within the various mobility subgroups. However, this conclusion must be tempered, because these descriptive findings may only explain differences that already existed prior to students' mobility.
ANOVAs revealed that there was a statistical difference at the p < .05 level on the CST in ELA and the California English Language Development Test (CELDT) between mean scores of stabile, mobile and highly mobile students. ANOVAs on the CST in general math and algebra revealed that there was a statistical difference in mean scores for general math for the non-EL population, but not for English learners. The ANOVAs confirmed that stabile students performed better on standardized tests in English language arts and general math than their more mobile peers. Stabile English learners performed better on the CELDT than their more mobile classmates.
Multiple regressions were conducted using a stepwise algorithm within a hierarchical method to determine which independent variables contributed to the models' ability to predict mobility and predict performance on the CST in ELA and mathematics. For the regressions predicting mobility, each model was found to be statistically significant, but representing a weak to moderately weak effect. According to the results of the regressions predicting performance on the CST in ELA and mathematics, mobility was found to be a significant predictor for the total sample and for English learners. Based on the analysis of the betas, variables representing prior achievement were the greatest predictors of future achievement on the CST, along with fundamental school attendance, English proficiency, gender, and parent education.
Discriminant function analysis was used to predict membership in one of two groups, "stabile" or "mobile" for the total sample and EL subpopulation. The discriminants were found to be significant and were able to classify 72.4% and 67.5% of original grouped cases correctly. This finding yields important information for schools within SAUSD, as it can be used to identify at-risk students and provide them the academic assistance and counseling necessary to moderate the negative effects of mobility.
This study supports previous research that mobility impacts student achievement in ELA and mathematics. This study also builds on what is known about the impact of mobility on English learners and English language development by finding that: (1) There is a disproportionate percentage of English learners within the mobile and highly mobile subgroups; (2) There is a significant difference between stabile, mobile and highly students on tests of English language proficiency; and (3) Being an English learner is a significant predictor of student mobility.