The supermatrix technique: A simple framework for hypothesis testing with missing data

被引：23

作者：

Lang, Kyle M. ^{[1
]}

Little, Todd D. ^{[2
]}

机构：

[1] Univ Kansas, Lawrence, KS 66045 USA

[2] Texas Tech Univ, Lubbock, TX 79409 USA

来源：

INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT | 2014年 / 38卷 / 05期

基金：

美国国家科学基金会;

关键词：

full information maximum likelihood; hypothesis testing; missing data; Monte Carlo simulation; multiple imputation; IMPUTATION; INFERENCE;

D O I：

10.1177/0165025413514326

中图分类号：

B844 [发展心理学（人类心理学）];

学科分类号：

040202 ;

摘要：

We present a new paradigm that allows simplified testing of multiparameter hypotheses in the presence of incomplete data. The proposed technique is a straight-forward procedure that combines the benefits of two powerful data analytic tools: multiple imputation and nested-model chi(2) difference testing. A Monte Carlo simulation study was conducted to assess the performance of the proposed technique. Full information maximum likelihood (FIML) and single regression imputation were included as comparison conditions against which the performance of the suggested technique was judged. The imputation-based conditions demonstrated much higher convergence rates than the FIML conditions. Delta chi(2) statistics derived from the proposed technique were more accurate than such statistics derived from both the FIML conditions and the regression imputation conditions. Limitations of the current work and suggestions for future directions are also addressed.

引用

页码：461 / 470

页数：10