Comparison of REML and MINQUE for Estimated Variance Components and Predicted Random Effects  

作　　者：Nan Nan Johnie N. Jenkins Jack C. McCarty Jixiang Wu Nan Nan;Johnie N. Jenkins;Jack C. McCarty;Jixiang Wu(Department of Mathematics and Statistics;South Dakota State University, Brookings, SD, USA;Crop Science Research Laboratory USDA-ARS, Mississippi State, MS, USA;Agronomy, Horticulture, and Plant Science Department, South Dakota State University, Brookings, SD, USA)

机构地区：[1]Department of Mathematics and Statistics South Dakota State University, Brookings, SD, USA [2]Crop Science Research Laboratory USDA-ARS, Mississippi State, MS, USA [3]Agronomy, Horticulture, and Plant Science Department, South Dakota State University, Brookings, SD, USA

出　　处：《Open Journal of Statistics》2016年第5期814-823,共11页统计学期刊（英文）

摘　　要：Linear mixed model (LMM) approaches have been widely applied in many areas of research data analysis because they offer great flexibility for different data structures and linear model systems. In this study, emphasis is placed on comparing the properties of two LMM approaches: restricted maximum likelihood (REML) and minimum norm quadratic unbiased estimation (MINQUE) with and without resampling techniques being included. Bias, testing power, Type I error, and computing time were compared between REML and MINQUE approaches with and without Jackknife technique based on 500 simulated data sets. Results showed that MINQUE and REML methods performed equally regarding bias, Type I error, and power. Jackknife-based MINQUE and REML greatly improved power compared to non-Jackknife based linear mixed model approaches. Results also showed that MINQUE is more time-saving compared to REML, especially with the use of resampling techniques and large data set analysis. Results from the actual cotton data analysis were in agreement with our simulated results. Therefore, Jackknife-based MINQUE approaches could be recommended to achieve desirable power with reduced time for a large data analysis and model simulations.Linear mixed model (LMM) approaches have been widely applied in many areas of research data analysis because they offer great flexibility for different data structures and linear model systems. In this study, emphasis is placed on comparing the properties of two LMM approaches: restricted maximum likelihood (REML) and minimum norm quadratic unbiased estimation (MINQUE) with and without resampling techniques being included. Bias, testing power, Type I error, and computing time were compared between REML and MINQUE approaches with and without Jackknife technique based on 500 simulated data sets. Results showed that MINQUE and REML methods performed equally regarding bias, Type I error, and power. Jackknife-based MINQUE and REML greatly improved power compared to non-Jackknife based linear mixed model approaches. Results also showed that MINQUE is more time-saving compared to REML, especially with the use of resampling techniques and large data set analysis. Results from the actual cotton data analysis were in agreement with our simulated results. Therefore, Jackknife-based MINQUE approaches could be recommended to achieve desirable power with reduced time for a large data analysis and model simulations.

关 键 词：Comparison of REML and MINQUE for Estimated Variance Components and Predicted Random Effects 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]