Work

Evaluation of single- and multilevel factor mixture model estimation

Year: 2007
Type: article
Abstract: Confirmatory factor analysis (CFA) models test the plausibility of latent \nconstructs hypothesized to account for relations among observed variables. CFA models \ncan be used to model both hierarchical data structures as a result of cluster sampling \ndesigns and investigate the plausibility of latent classes or unobserved classes of \nindividuals. Recent research has suggested preliminary evidence on the accuracy of \ntypical fit indices (AIC, BIC, aBIC, LMR aLRT) among various single-level latent class \nmodels, multilevel latent class models that correct for biased standard error estimates as a \nresult of nested data, and growth mixture models (Clark & Muthén, 2007; Nylund et al., \n2006, Tofighi & Enders, 2006). But few, if any, researchers have studied the accuracy of \nthe fit indices in multilevel factor mixture models. The purpose of this study was to \nextend the literature in this less researched area and assess the performance of typical fit \nindices used to compare factor mixture models. Class separation, intraclass correlation, \nand between-cluster sample size were manipulated to emulate realistic typical research \nconditions. \n The proportion of times out of a possible 100 replications in which each of the \nAIC, BIC, aBIC, and LMR aLRT led to selection of the correctly specified model among \nother mis-specified models was recorded. Results for data generated to fit one-class \nmodels indicated that the BIC and aBIC outperformed the AIC and the LMR aLRT was \nnonsignificant nearly 100% of the time, supporting the correct one-class model. \nPerformance of all fit indices was, however, poor when data were generated to originate \nfrom two-class models. The AIC and LMR aLRT tended to perform better than the BIC \nand aBIC, although accuracy of all fit indices increased as a function of increasing class \nseparation and between-cluster sample size. Implications and recommendations regarding \noptimal fit indices under various conditions are reported. It is hoped that the current \nresearch has provided initial evidence of conditions in which the various fit indices are \nmore likely to model the correct number of latent classes in multilevel data.
Author: Shane Allua
Language: English
Open Access status: green
OpenAlex ID: W1694539035
FWCI: 0.389
Cites: 86
Cited by: 7
Related to: 10