多层线性模型在纵向研究中的运用

The Application Of Multilevel Model In Longitudinal Research

纵向研究中传统统计技术主要是重复测量的方差分析和多元回归分析,但是这两种技术存在一些局限性,不能合理而充分地解释纵向研究资料。近年来出现的多层线性模型能够更有效地利用纵向数据。为促进多层线性模型在纵向研究中的运用,简要论述了传统统计技术的局限,介绍了多层线性模型的原理及其在纵向研究中的作用,以一组模拟数据为例对多层线性模型中各种参数在纵向研究中的含义进行了详细讨论,对于纵向研究中运用多层线性模型时须注意的问题提出了建议。

Some limits exist in traditional longitudinal analysis technique such as repeated ANOVA and multivariate regression. The emergence of multilevel analy sis improved the power in longitudinal data analysis. The limits of traditional methods and the advantage of multilevel analysis were introduced. Meanings of pa rameters of multilevel model in longitudinal research were discussed and an exam ple data were used to illuminate these meanings. Some suggestions about future a pplication of multilevel model in longitudinal research were put forward.