175 – SPEED: Recent Advance of Statistical Methods in Biometrics
The Delta Garden Study: A Quasi-Experimental, Cross-Sectional, Nested, Pair-Matched Design with Zero-Inflated Endpoints
Page Moore, PhD
University of Arkansas for Medical Sciences
Amy Schrader, M.S.
University of Arkansas for Medical Sciences
Judith L. Weber, PhD, RD
Arkansas Children’s Hospital Research Institute
The Delta Garden Study is a science-based school garden intervention designed to increase fruit and vegetable (FV) intakes and minutes of moderate to vigorous physical activity (PA), as well as to improve academic achievement and reduce social risk behaviors, in middle school students in Arkansas. The study used a quasi-experimental, cross-sectional, nested, pair-matched design utilizing 6 intervention (school garden) schools and 6 control (no school garden) schools, including over 2,000 6th, 7th and 8th grade students nested within 4-6 science teachers per school. Observed as continuous outcomes, the two primary endpoints were zero-inflated and were collected across multiple time-points. We present the design and analysis of this quasi-experimental, cross-sectional, nested, pair-matched design and the handling of longitudinally collected zero-inflated outcomes.