A two‐stage estimation of structural equation models with continuous and polytomous variables

Lee, Sik‐Yum; Poon, Wai‐Yin; Bentler, Peter M.

doi:10.1111/j.2044-8317.1995.tb01067.x

Cited by 215 publications

(166 citation statements)

References 11 publications

Supporting

Mentioning

151

Contrasting

Unclassified

Order By: Relevance

“…Tests of multivariate normality (Mardia, 1985), as well as standard solutions to overcome these problems, such as the use of the ADF estimator (Browne, 1984) and/or the use of categorized data procedures (e.g. Lee et al, 1995), are applicable only with listwise deletion treatment of missing values. Thus, using FIML may trade off bias caused by suboptimal missing data treatment against bias due to nonnormality of the data.…”

Section: Statistical Modellingmentioning

confidence: 99%

Development of Life Satisfaction in Old Age: Another View on the "Paradox''

Schilling

2006

Soc Indic Res

115

View full text Add to dashboard Cite

ABSTRACT. Empirical evidence of no age-related decline in life satisfaction (LS) in old age contrasts with frequently observed declines in the objective quality of elder people's lives and has therefore been labelled a ''paradox'' and interpreted in terms of stability of LS in the respective gerontological discussion. However, as this evidence was mainly derived from cross-sectional age group comparisons, it does neither clearly indicate the absence of age-related mean level change, nor intra-individual stability of LS. Thus, the development of LS in middle and late adulthood was analysed longitudinally by using data from the German Socio-Economic Panel. Based on single item measures of LS across 16 repeated panel waves with one per annum (1984)(1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995)(1996)(1997)(1998)(1999), autoregressive (quasi-Markov) structural equation models were used to estimate true score variances and intra-individual true score stability in oneyear intervals. Research questions concerned (a) ''monotonic'' stability and variance in a subsample of old respondents (born before 1925) as compared to the total sample and (b) change in stability and variances in old age. Results indicate high ''monotonic'' true score stability of LS over the whole adult life span, whereas mean levels declined slightly in old age. No striking evidence for age-related changes in variance or stability was found.

show abstract

Section: Statistical Modellingmentioning

confidence: 99%

Development of Life Satisfaction in Old Age: Another View on the "Paradox''

Schilling

2006

Soc Indic Res

115

View full text Add to dashboard Cite

show abstract

“…No special purpose computer program is required. However, the two most popular programs for covariance structure analysis, LISREL (Jrreskog & Srrbom, 1993) and EQS (Bentler, 1995) cannot currently be used to fit Thurstonian ranking models since they implement correlation structure analysis for binary dependent variables, but not mean and correlation structure analysis (see Lee, Poon & Bentler, 1995;Jrreskog, 1994) required for Thurstonian modeling of ranking data. LISCOMP (Muthrn, 1987) performs mean and correlation structure analysis of binary dichotomous variables using Muthrn's (1978) estimator.…”

Section: Discussionmentioning

confidence: 99%

“…Just as Christofferson's estimator for the normal ogive model can be made computationally more efficient by making it two-stage (Muthtn, 1978), Chan and Bentler (1998) proposed a two-stage limited information estimator for Thurstonian models that uses information from trinary rankings. In the first stage, Chan and Bentler estimate an unrestricted Thurstonian model using either Lee, Poon and Bentler's (1995) partitioned maximum likelihood approach, or a direct weighted minimum distance approach as in Brady (1989). In a second stage, restricted Thurstonian models are estimated from the first stage results using a weighted minimum distance procedure.…”

Section: Introductionmentioning

confidence: 99%

Thurstonian modeling of ranking data via mean and covariance structure analysis

Maydeu-Olivares

1999

Psychometrika

View full text Add to dashboard Cite

Although Thurstonian models provide an attractive representation of choice behavior, they have not been extensively used in ranking applications since only recently efficient estimation methods for these models have been developed. These, however, require the use of special-purpose estimation programs, which limits their applicability. Here we introduce a formulation of Thurstonian ranking models that turns an idiosyncratic estimation problem into an estimation problem involving mean and covariance structures with dichotomous indicators. Well-known standard solutions for the latter can be readily applied to this specific problem, and as a result any Thurstonian model for ranking data can be fitted using existing general purpose software for mean and covariance structure analysis. Although the most popular programs for covafiance structure analysis (e.g., LISREL and EQS) cannot be presently used to estimate Thurstonian ranking models, other programs such as MECOSA already exist that can be straightforwardly used to estimate these models.

show abstract

“…This is also the case for multi-sample models. To solve this problem, we use the common method (see, for example, Lee et al, 1995;Shi and Lee, 1998) of fixing some thresholds at preassigned values. For convenience, we assume that the positions of the fixed elements are the same for each group.…”

Section: Model Descriptionmentioning

confidence: 99%

Bayesian Analysis of Linear and Nonlinear Latent Variable Models with Fixed Covariate and Ordered Categorical Data

Thanoon

Adnan

2016

Pak.j.stat.oper.res.

View full text Add to dashboard Cite

In this paper, ordered categorical variables are used to compare between linear with covariate and nonlinear interactions of covariates and latent variables in Bayesian structural equation models. Gibbs sampling method is applied for estimation and model comparison. Hidden continuous normal distribution (censored normal distribution) is used to handle the problem of ordered categorical data. Statistical inferences, which involve estimation of parameters and their standard deviations, and residuals analyses for testing the selected model, are discussed. The proposed procedure is illustrated by a real data. Analyses are done by using OpenBUGS program.

show abstract

A two‐stage estimation of structural equation models with continuous and polytomous variables

Cited by 215 publications

References 11 publications

Development of Life Satisfaction in Old Age: Another View on the "Paradox''

Development of Life Satisfaction in Old Age: Another View on the "Paradox''

Thurstonian modeling of ranking data via mean and covariance structure analysis

Bayesian Analysis of Linear and Nonlinear Latent Variable Models with Fixed Covariate and Ordered Categorical Data

Contact Info

Product

Resources

About