Student Mobility Patterns across Virginia public schools using VLDS
Author: Isabel Bradburn, Virginia Tech
October 25, 2016
Isabel S. Bradburn, Child Development Center for Learning and Research, Virginia Tech
Many factors affect school children’s learning, and states and localities are increasingly focusing attention on how to best deliver education to a wide variety of students. Administrative data systems like the VLDS offer policymakers an efficient way to identify and then track trends over time associated with differences in scholastic achievement that may help benchmark progress or suggest when or where to enhance educational supports. In this blog, we describe how researchers at Virginia Tech, working together with the Virginia Department of Education, are using the VLDS to look at student mobility patterns across Virginia public schools.
By “mobility,” we mean students transferring between Virginia public schools or divisions. We focused on student mobility because research supports the commonsense notion that students who transfer school frequently are more at-risk for adverse academic outcomes. One meta-analysis (a statistical summary of many study findings) estimated a 3-4 month delay for every school transfer. But just how widespread is student mobility? Early national surveys indicated about 11% of students transfer once during the academic year, while 13% change schools four or more times by high school; single school district annual “turnover” rates vary widely, depending on district and grade. To assess mobility rates and patterns in Virginia, we are examining school enrollment records over time across elementary school - that is, from kindergarten through fifth grade. We started with these grades because research indicates that frequent changes in the earliest years of formal schooling may have the longest-term negative effects.
Results from the first wave of analyses with a single cohort show a fair degree of stability across elementary school, although mobility rates vary across different school divisions. For example, 84% of students who started kindergarten in 2008 were in the public school system by the end of academic year 2013-2014, which for most students represented the end of elementary school (fifth grade). A slim majority (53%) remained in the same school across the entire time period, while about one fifth (19.6%) transferred two or more times. Transferring between school years was more typical (40%) than transferring during the school year (18%) (and some students did both). About seven percent of students in this cohort left and then returned to the public schools before sixth grade.
Sometimes students change schools because of the way schools are structured, such as moving from an elementary to a middle school. We found that school transfer rates were higher between some elementary grades than others. Changing schools between second and third grades was the most common transfer pattern, followed by transfer between fourth and fifth grades. Of the 1302 Commonwealth public schools serving elementary grades in 2008-2014, the great majority offered prekindergarten through grade 5 within a single school. The next most common schooling structure offered prekindergarten through second grade in one school, followed by grades 3 through 5 in a separate school. This normative transition from early to later elementary school present in some school divisions may be a major reason for the elevated transfer rates between second and third grades. An expected, normative transition that students share with their peers may be associated with very different academic outcomes compared to changing schools due to other reasons.
The next step will be to examine mobility patterns across additional cohorts and the extent to which student mobility is associated with student achievement. By documenting rates of student mobility over multiple cohorts, we should get a better sense for what might be considered a “typical” level of transfer for the state and across divisions, when spikes or lulls may occur, and what types or degrees of mobility may be related to particular student and school educational outcomes in Virginia. Charting trends and establishing multi-year “typical” rates can also provide a yardstick by which to measure “true” change. The VLDS is a powerful tool capable of addressing these and related questions with greater precision and accuracy than much previous research that had to rely on retrospective student or parent reports or single-year samples.
Note: This research is supported by the National Science Foundation.