The small sample performance of estimators of the standard errors of structural equation models
2013 (English)In: Journal of Statistical Computation and Simulation, ISSN 0094-9655, E-ISSN 1563-5163, Vol. 83, no 3, 458-471 p.Article in journal (Refereed) Published
n this paper, we compare five asymptotically, under a correctly specified likelihood, equivalent estimators of the standard errors for parameters in structuralequation models. The estimators are evaluated under different conditions regarding (i)sample size, varying between N=50 and 3200, (ii) distributional assumption of the latent variables and the disturbance terms, namely normal, and heavy tailed (t), and (iii) thecomplexity of the model. For the assessment of the five estimators we use overallperformance, relative bias, MSE and coverage of confidence intervals. The analysis reveals substantial differences in the performance of the five asymptotically equalestimators. Most diversity was found for t distributed, i.e. heavy tailed, data.
Place, publisher, year, edition, pages
2013. Vol. 83, no 3, 458-471 p.
information matrix, sandwich estimator Monte Carlo simulation, estimators of standard errors, small sample sizes, structural equation modelling, SEM
Probability Theory and Statistics
Research subject Statistics
IdentifiersURN: urn:nbn:se:uu:diva-160763DOI: 10.1080/00949655.2011.616205ISI: 000315188300005OAI: oai:DiVA.org:uu-160763DiVA: diva2:452667