Managing Clustered Data Using Hierarchical Linear Modeling

详细信息	查看全文 | 推荐本文 |

• 作者：Russell T. Warne ; PhD¹ ; ^&dagger ; ; ^{rwarne@uvu.edu} ; Yan Li ; MS² ; ³ ; E. Lisako J. McKyer ; PhD ; MPH³ ; ⁴ ; ⁵ ; Rachel Condie ; BS³ ; ⁶ ; Cassandra S. Diep ; MS³ ; ⁴ ; Peter S. Murano ; PhD³ ; ⁷ ; ⁸
• 关键词：hierarchical linear modeling ; multilevel models ; WIC program ; survey research ; nutrition behavior
• 刊名：Journal of Nutrition Education and Behavior
• 出版年：2012
• 期刊代码：P99_14994046
• 类别：med
• 出版时间：May-June, 2012
• 卷：44
• 期：3
• 页码：271-277
• 文件大小：910 K

摘要

Researchers in nutrition research often use cluster or multistage sampling to gather participants for their studies. These sampling methods often produce violations of the assumption of data independence that most traditional statistics share. Hierarchical linear modeling is a statistical method that can overcome violations of the independence assumption and lead to correct analysis of data, yet it is rarely used in nutrition research. The purpose of this viewpoint is to illustrate the benefits of hierarchical linear modeling within a nutrition research context.