Running head: Test For Nonnested Models A Statistical Test For Comparing Nonnested Covariance Structure Models
نویسندگان
چکیده
While statistical procedures are well known for comparing hierarchically related (nested) covariance structure models, statistical tests for comparing nonhierarchically related (nonnested) models have proven more elusive. While isolated attempts have been made, none exists within the commonly used maximum likelihood estimation framework, thereby compromising these methods' accessibility and general applicability. The current work builds on a distance measure originally proposed by C. Rao (1945; 1949), and its application to distances between covariance structure models (A. Kumar and S. Sharma, 1999), thereby proposing a method for conducting a statistical test of such distances in order to assess formally the distinctness between modelsnested or nonnested. An illustration is presented, and simulation evidence is provided to validate the performance of the proposed method. Two appendixes contain an illustration of the model for data generation and a program to compute distances between covariance matrices. (Contains 1 table, 2 figures, and 28 references.) (Author/SLD) Reproductions supplied by EDRS are the best that can be made from the original document Test For Nonnested Models 1 Running head: Test For Nonnested Models A Statistical Test For Comparing Nonnested Covariance Structure Models Roy Levy and Gregory R. Hancock University of Maryland, College Park [Paper presented at the annual meeting of the American Educational Research Association, Chicago, IL, April 2003] U.S. DEPARTMENT OF EDUCATION Office of Educational Research and Improvement EDUCATIONAL RESOURCES INFORMATION CENTER (ERIC) GI This document has been reproduced as received from the person or organization originating it. Minor changes have been made to improve reproduction quality. Points of view or opinions stated in this document do not necessarily represent official OERI position or policy. PERMISSION TO REPRODUCE AND DISSEMINATE THIS MATERIAL HAS BEEN GRANTED BY
منابع مشابه
Statistical Tests for Comparing Possibly Misspecified and Nonnested Models.
Model selection criteria (MSC) involves selecting the model with the best estimated goodness-of-fit to the data generating process. Following the method of Vuong (1989), a large sample Model Selection Test (MST), is introduced that can be used in conjunction with most existing MSC procedures to decide if the estimated goodness-of-fit for one model is significantly different from the estimated g...
متن کاملSize Distortion and Modication of Classical Vuong Tests1
In this paper, I examine the nite sample properties of the Vuong test (Vuong (1989)) for nonnested model comparison. I nd that both the one-step and the two-step classical Vuong tests over-reject the null hypothesis when the null hypothesis is true. The over-rejection can be large in certain scenarios. A simple one-step modi ed test is proposed. The modi ed test applies to both overlapping mo...
متن کاملAn Encompassing Approach to Conditional Mean Tests with Applications to Testing Nonnested Hypotheses
A general class of tests designed to detect conditional mean misspecification for cross section or time series applications is proposed. The tests are derived from a particular application of the encompassing principle. The resulting conditional mean encompassing (CME) tests contain as special cases a version of the Lagrange Multiplier test for nested models, a new test in the presence of nonne...
متن کاملMarginal modeling of nonnested multilevel data using standard software.
Epidemiologic data are often clustered within multiple levels that may not be nested within each other. Generalized estimating equations are commonly used to adjust for correlation among observations within clusters when fitting regression models; however, standard software does not currently accommodate nonnested clusters. This paper introduces a simple generalized estimating equation strategy...
متن کاملOriginal Contribution Marginal Modeling of Nonnested Multilevel Data using Standard Software
Epidemiologic data are often clustered within multiple levels that may not be nested within each other. Generalized estimating equations are commonly used to adjust for correlation among observations within clusters when fitting regression models; however, standard software does not currently accommodate nonnested clusters. This paper introduces a simple generalized estimating equation strategy...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2012