WITHDRAWN: Evaluating School-Based Interventions Using Quantile Regression Models of Longitudinal Growth Trajectories: An Application Among NYC Public School Elementary- and Middle-School Students

• Activity Number: 34
• Type: Contributed
• Date/Time: Sunday, August 3, 2014 : 2:00 PM to 3:50 PM
• Sponsor: Health Policy Statistics Section
• Author(s): Kevin Konty and Stuart Sweeney and Sophia Day
• Companies: and University of California and NYC DOHMH
• Keywords: Quantile regression ; public health ; obesity ; education ; policy evaluation ; administrative data

Abstract

As part of the NYC FITNESSGRAM curriculum implemented in 2005, NYC public schools keep annual records of student body mass indices (BMIs); such school-based collection of student height and weight data has become increasingly common. As such, large longitudinal datasets containing individual growth trajectories are now available. A use of this data is for the evaluation of programs or interventions implemented within schools. Current approaches such as difference-in-difference models focus on mean changes in BMI across all students and often failto use the longitudinal nature of the data, while longitudinal models that explicitly address obesity status suffer from a loss of information by simply using a threshold as an indicator of change. We describe an implementation of quantile regression models of longitudinal growth trajectories that incorporate classroom-, school-, and city-wide interventions to NYC FITNESSGRAM data focusing on practical statistical considerations. These models characterize the impact of these interventions on the tails of the student BMI distributions. We will focus on an application to the Breakfast-in-Classroom program and its impact on BMI.