Murray D M, Phillips G A, Bimbaum A S, Lytle L A
Department of Psychology, University of Memphis, Tennessee 38152-3230, USA.
Health Educ Behav. 2001 Dec;28(6):666-79. doi: 10.1177/109019810102800602.
This article presents the first estimates of school-level intraclass correlation for dietary measures based on data from the Teens Eating for Energy and Nutrition at School study. This study involves 3,878 seventh graders from 16 middle schools from Minneapolis-St. Paul, Minnesota. The sample was 66.8% White, 11.2% Black, and 7.0% Asian; 48.8% of the sample was female. Typical fruit and vegetable intake was assessed with a modified version of the Behavior Risk Factor Surveillance System questionnaire. Twenty-four-hour dietary recalls were conducted by nutritionists using the Minnesota Nutrition Data System. Mixed-model regression methods were used to estimate variance components for school and residual error, both before and after adjustment for demographic factors. School-level intraclass correlations were large enough, if ignored, to substantially inflate the Type I error rate in an analysis of treatment effects. The authors show how to use the estimates to determine sample size requirements for future studies.
本文基于“青少年在校饮食与营养”研究的数据，首次给出了学校层面饮食测量的组内相关估计值。该研究涉及明尼苏达州明尼阿波利斯 - 圣保罗市16所中学的3878名七年级学生。样本中白人占66.8%，黑人占11.2%，亚裔占7.0%；48.8%的样本为女性。使用行为风险因素监测系统问卷的修订版评估了典型的水果和蔬菜摄入量。营养学家使用明尼苏达营养数据系统进行了24小时饮食回顾。在对人口统计学因素进行调整前后，均使用混合模型回归方法估计学校和残差误差的方差成分。学校层面的组内相关性足够大，如果忽略不计，会在治疗效果分析中大幅夸大I型错误率。作者展示了如何使用这些估计值来确定未来研究的样本量要求。
