用于估计多级测量模型参数的软件程序演示。

Demonstration of software programs for estimating multilevel measurement model parameters.

作者信息

Roberts J Kyle, Herrington Rich

机构信息

Baylor College of Medicine, One Baylor Plaza, BCM 155, Houston, TX 77030, USA.

出版信息

J Appl Meas. 2005;6(3):255-72.

PMID:15942070

Abstract

A brief overview of the relevant parameterizations of the Rasch measurement model will first be provided. Next, the use of five different hierarchical linear modeling software programs (SAS, MLWIN, S-PLUS, R and HLM) using a heuristic data set will be demonstrated. For each program, researchers will be offered: 1) instructions describing the way to set up data sets, as well as 2) directions for running each program, 3) guidelines to assist with the appropriate interpretation of the output. Last, testing of the local independence assumption will be discussed. All data and code for each of the models run in this analysis may be downloaded from the website: http://www.hlm-online.com/papers/mmmpaper.htm/.

摘要

首先将简要概述拉施测量模型的相关参数化。接下来，将演示使用启发式数据集的五种不同分层线性建模软件程序（SAS、MLWIN、S-PLUS、R和HLM）的使用方法。对于每个程序，将为研究人员提供：1）描述设置数据集方法的说明，以及2）运行每个程序的指导，3）有助于正确解释输出的指南。最后，将讨论局部独立性假设的检验。本次分析中运行的每个模型的所有数据和代码均可从网站：http://www.hlm-online.com/papers/mmmpaper.htm/下载。