Submitted to: International Society for Behavioral Nutrition and Physical Activity
Publication Type: Abstract Only
Publication Acceptance Date: 4/20/2005
Publication Date: 6/16/2005
Citation: Gossett, J., Simpson, P., Jo, C., Yadrick, K., Champagne, C., Bogle, M.L. 2005. A picture is worth a thousand words for describing eating patterns [abstract]. Proceedings of International Society for Behavioral Nutrition and Physical Activity. p. 161. Interpretive Summary:
Technical Abstract: Purpose: To show how new graphical capabilities in available software impact investigation of interrelationships of nutrients and foods. Background: A summary table of many nutrients and foods of interest for several different demographic groups can be difficult to digest. The information in the new graphical representations possible today is much easier to swallow. Edit checks, checks of model assumptions and displays of eating habits can be concisely summarized in a graph that is easy to understand. Methods: Graphs reveal data properties and can make large data sets comprehensible. Since Nutrition data is multivariate, we need creative ways to display data simultaneously so we can see interrelationships. In a trellis plot we display the Healthy Eating Indices in a survey and the percentage that meet the Dietary Reference Intakes (DRI) for 12 different nutrients and overlay the results of different subgroups (black-vs.-white, low-vs-medium-vs-high income, male-vs.-female, and so forth). We also show how trellis plots help to identify data entry errors and outliers in a food store survey. Conclusions: Although modeling allows adjustment for other variables, checking assumptions of modeling and understanding relationships with the aid of graphs is an essential component to truly describing eating patterns.