From Cross-Sectional to Cohort Data: The Creation of the Act 1220 Evaluation Longitudinal Databases

• Activity Number: 533
• Type: Contributed
• Date/Time: Wednesday, August 4, 2010 : 10:30 AM to 12:20 PM
• Sponsor: Section for Statistical Programmers and Analysts
• Author(s): Melanie E. Goodell*+ and Zoran Bursac and Clinton Heath Gauss and Martha M. Phillips
• Companies: University of Arkansas for Medical Sciences and University of Arkansas for Medical Sciences and University of Arkansas for Medical Sciences and University of Arkansas for Medical Sciences
• Keywords: childhood obesity ; longitudinal studies ; data management ; evaluation

Abstract

Arkansas Act 1220 of 2003 was legislation aimed at halting the rise in childhood obesity rates. Researchers at the UAMS College of Public Health began a comprehensive evaluation of the legislation's impact in early 2004. The team has surveyed all principals and superintendents in the state annually, leading to the creation of yearly cross-sectional data sets for schools and districts. While these unlinked databases have answered past research questions, linking the data sets across years allows us to study change in an individual school's policies. The process of linking data from years 1 through 6 of the evaluation had to account for district consolidations, school closures, new schools, and changes in school structure. This paper will explain the procedures used to create the longitudinal databases and discuss the challenges inherent in working with a dynamic population of schools.