Abstract
A 2-year supermarket intervention in 20 pair-matched stores tested whether a nutrition education program could produce changes in food purchasing behavior. Food sales data for 2 intervention years were compared with baseline year data. The evaluation presented challenges related to the quasi-experimental study design and to the high variability and auto- and cross-sectional correlations in the data. A complex time series model was used to take these factors into account and to test for intervention effects.
Get full access to this article
View all access options for this article.
