Reliable and More Powerful Methods for Power Analysis in Structural Equation Modeling

Autor/inn/en	Yuan, Ke-Hai; Zhang, Zhiyong; Zhao, Yanyun
Titel	Reliable and More Powerful Methods for Power Analysis in Structural Equation Modeling
Quelle	24 (2017) 3, S.315-330 (17 Seiten)Infoseite zur Zeitschrift PDF als Volltext (1); PDF als Volltext kostenfreie Datei (2) Verfügbarkeit
Zusatzinformation	Weitere Informationen
Sprache	englisch
Dokumenttyp	gedruckt; online; Zeitschriftenaufsatz
ISSN	1070-5511
DOI	10.1080/10705511.2016.1276836
Schlagwörter	Statistical Analysis; Evaluation Methods; Structural Equation Models; Reliability; Monte Carlo Methods; Effect Size; Robustness (Statistics); Error of Measurement; Maximum Likelihood Statistics; Statistical Distributions; Sample Size; Factor Analysis + Suchen Sie Ihr Suchwort? Statistische Analyse; Reliabilität; Monte-Carlo-Methode; Widerstandsfähigkeit; Messfehler; Wahrscheinlichkeitsverteilung; Faktorenanalyse
Abstract	The normal-distribution-based likelihood ratio statistic T[subscript ml] = nF[subscript ml] is widely used for power analysis in structural Equation modeling (SEM). In such an analysis, power and sample size are computed by assuming that T[subscript ml] follows a central chi-square distribution under H[subscript 0] and a noncentral chi-square distribution under H[subscript a]. However, with either violation of normality or not a large enough sample size, both empirical and analytical results indicate that the chi-square distribution assumptions are not realistic and consequently methods of power analysis based on such assumptions are not valid. This article describes a Monte Carlo (MC) method for power analysis. A measure of effect size for characterizing the power property of different rescaled statistics is also provided. Robust methods are proposed to increase the power of T[subscript ml] and other statistics. Simulation results show that the MC method reliably controls Type I errors and robust estimation methods effectively increase the power, and their combination is thus recommended for conducting power analysis in SEM. (As Provided).
Erfasst von	ERIC (Education Resources Information Center), Washington, DC
Update	2020/1/01