Variance component testing in multilevel models

Johannes Berkhof; Tom A.B. Snijders

doi:https://doi.org/10.3102/10769986026002133

Variance component testing in multilevel models

Johannes Berkhof, Tom A.B. Snijders

Research output: Contribution to journal › Article › Academic › peer-review

65 Citations (Scopus)

Abstract

Available variance component tests are reviewed and three new score tests are presented. In the first score test, the asymptotic normal distribution of the test statistic is used as a reference distribution. In the other two score tests, a Satterthwaite approximation is used for the null distribution of the test statistic. We evaluate the performance of the score tests and other available tests by means of a Monte Carlo study. The new tests are computationally relatively cheap and have good power properties.

Original language	English
Pages (from-to)	133-152
Number of pages	20
Journal	JOURNAL OF EDUCATIONAL AND BEHAVIORAL STATISTICS
Volume	26
Issue number	2
DOIs	https://doi.org/10.3102/10769986026002133
Publication status	Published - 2001

Keywords

Monte Carlo study
Multilevel models
Random coefficients
Score tests
Variance components

Access to Document

https://doi.org/10.3102/10769986026002133

Cite this

@article{c984dfe79713401db2e83159fa459546,

title = "Variance component testing in multilevel models",

abstract = "Available variance component tests are reviewed and three new score tests are presented. In the first score test, the asymptotic normal distribution of the test statistic is used as a reference distribution. In the other two score tests, a Satterthwaite approximation is used for the null distribution of the test statistic. We evaluate the performance of the score tests and other available tests by means of a Monte Carlo study. The new tests are computationally relatively cheap and have good power properties.",

keywords = "Monte Carlo study, Multilevel models, Random coefficients, Score tests, Variance components",

author = "Johannes Berkhof and Snijders, {Tom A.B.}",

year = "2001",

doi = "https://doi.org/10.3102/10769986026002133",

language = "English",

volume = "26",

pages = "133--152",

journal = "JOURNAL OF EDUCATIONAL AND BEHAVIORAL STATISTICS",

issn = "1076-9986",

publisher = "SAGE Publications Inc.",

number = "2",

}

TY - JOUR

T1 - Variance component testing in multilevel models

AU - Berkhof, Johannes

AU - Snijders, Tom A.B.

PY - 2001

Y1 - 2001

N2 - Available variance component tests are reviewed and three new score tests are presented. In the first score test, the asymptotic normal distribution of the test statistic is used as a reference distribution. In the other two score tests, a Satterthwaite approximation is used for the null distribution of the test statistic. We evaluate the performance of the score tests and other available tests by means of a Monte Carlo study. The new tests are computationally relatively cheap and have good power properties.

AB - Available variance component tests are reviewed and three new score tests are presented. In the first score test, the asymptotic normal distribution of the test statistic is used as a reference distribution. In the other two score tests, a Satterthwaite approximation is used for the null distribution of the test statistic. We evaluate the performance of the score tests and other available tests by means of a Monte Carlo study. The new tests are computationally relatively cheap and have good power properties.

KW - Monte Carlo study

KW - Multilevel models

KW - Random coefficients

KW - Score tests

KW - Variance components

UR - http://www.scopus.com/inward/record.url?scp=0035636276&partnerID=8YFLogxK

U2 - https://doi.org/10.3102/10769986026002133

DO - https://doi.org/10.3102/10769986026002133

M3 - Article

SN - 1076-9986

VL - 26

SP - 133

EP - 152

JO - JOURNAL OF EDUCATIONAL AND BEHAVIORAL STATISTICS

JF - JOURNAL OF EDUCATIONAL AND BEHAVIORAL STATISTICS

IS - 2

ER -

Variance component testing in multilevel models

Abstract

Keywords

Access to Document

Other files and links

Cite this