Evaluating estimation methods for ordinal data in structural equation modeling

This study examined the performance of two alternative estimation approaches in structural equation modeling for ordinal data under different levels of model misspecification, score skewness, sample size, and model size. Both approaches involve analyzing a polychoric correlation matrix as well as adjusting standard error estimates and model chi-squared, but one estimates model parameters with maximum likelihood and the other with robust weighted least-squared. Relative bias in parameter estimates and standard error estimates, Type I error rate, and empirical power of the model test, where appropriate, were evaluated through Monte Carlo simulations. These alternative approaches generally provided unbiased parameter estimates when the model was correctly specified. They also provided unbiased standard error estimates and adequate Type I error control in general unless sample size was small and the measured variables were moderately skewed. Differences between the methods in convergence problems and the evaluation criteria, especially under small sample and skewed variable conditions, were discussed.     

Estimating intercoder reliability: a structural equation modeling approach

Intercoder reliability is usually estimated with a summary index, and yet the limitations concerning the indexing approach have been well noted. This study critically reviewed all the existing major modeling approaches to estimating intercoder reliability, and empirically tested and further compared these approaches. It was found that latent variable modeling, also called the second-generation SEM, generally perform better than log-linear modeling, and is able to explain the paradox haunting some indices, and to spot the sources of disagreement among coders. Implications were discussed at last.     

Structural equation modeling and the latent linearity hypothesis in social and behavioral research

The issue of sensitivity of the structural equation modeling (SEM) methodology to violations of the underlying hypothesis of linear latent relationships is the focus of this paper. The identity of overall goodness-of-fit indices of an initially considered linear latent pattern model and of an equivalent model not making this assumption exemplifies the lack of routinely available global means within the methodology to evaluate the linearity assumption. It is next focused on the sensitivity of SEM to violations of presumed linearity for a general, nonlinear pattern of true relationship. The results of a simulation study are then presented which demonstrate that latent correlations and percentage explained variance as well as parameter standard errors and model residuals can provide critical information about violation of latent linearity, and should therefore also be focused on when examining departures from linear relationships at the latent level in applications of the SEM methodology in social and behavioral research.     

Robustness issues in structural equation modeling: a review of recent developments

In structural equation modeling the statistician needs assumptions inorder (1) to guarantee that the estimates are consistent for the parameters of interest, and (2) to evaluate precision of the estimates and significance level of test statistics. With respect to purpose (1), the typical type of analyses (ML and WLS) are robust against violation of distributional assumptions; i.e., estimates remain consistent or any type of WLS analysis and distribution of z. (It should be noted, however, that (1) is sensitive to structural misspecification.) A typical assumption used for purpose (2), is the assumption that the vector z of observable follows a multivariate normal distribution.In relation to purpose (2), distributional misspecification may have consequences for efficiency, as well as power of test statistics (see Satorra, 1989a); that is, some estimation methods may bemore precise than others for a given specific distribution of z. For instance, ADF-WLS is asymptotically optimal under a variety of distributions of z, while the asymptotic optimality of NT-WLS may be lost when the data is non-normal     

Semi-parametric estimation of simultaneous equations with limited dependent variables: A case study of female labour supply

We propose a general strategy to estimate semi-parametrically simultaneous equations with limited dependent variables. First, each reduced form (RF) is estimated with various semi-parametric methods. Second, the specification of each RF is tested to select an appropriate method. Third, the structural form (SF) equations are estimated using minimum distance methods and the restrictions among the SF and RF parameters. A case study of female labour supply is presented.     

Hierarchical analysis: Classification with ordinal object dissimilarities

Summary Lerman [1970] has demonstrated, that the dissimilarity indices normally used in data analysis are identical up to strictly monotone transformationsf:R ₊R ₊ if the data are nominal and each set of attribute scores is finite.In that case he proposes to use a preorder between pairs of objects to express similarity or dissimilarity, in order to avoid inconsistent classification results that might occur, if clustering schemes which are not monotone invariant are applied to a quantitative index. Here it is shown, how a hierarchy on the objects can be calculated if such a preorder relation is given.     

On finite sample properties of alternative estimators of coefficients in a structural equation with many instruments

We compare four different estimation methods for the coefficients of a linear structural equation with instrumental variables. As the classical methods we consider the limited information maximum likelihood (LIML) estimator and the two-stage least squares (TSLS) estimator, and as the semi-parametric estimation methods we consider the maximum empirical likelihood (MEL) estimator and the generalized method of moments (GMM) (or the estimating equation) estimator. Tables and figures of the distribution functions of four estimators are given for enough values of the parameters to cover most linear models of interest and we include some heteroscedastic cases and nonlinear cases. We have found that the LIML estimator has good performance in terms of the bounded loss functions and probabilities when the number of instruments is large, that is, the micro-econometric models with “many instruments” in the terminology of recent econometric literature.     

Building strategic alliances: a case study

It takes much coordination, communication, patience, and persistence to build and manage supplier partnerships in order to lower total costs, reduce lead times, improve the quality of products, optimize inventories, and streamline administrative process. In the absence of the latest technology, it takes talented and enthusiastic team members who take pride in their work, who can think progressively, who can generate optimism in any situation, and who can maintain grace under pressure in order to be successful.     

The impact of non-normality,sample size and estimation technique on goodness-of-fit measures in structural equation modeling: evidence from ten empirical models of travel behavior

Ten empirical models of travel behavior are used to measure the variability of structural equation model goodness-of-fit as a function of sample size, multivariate kurtosis, and estimation technique. The estimation techniques are maximum likelihood, asymptotic distribution free, bootstrapping, and the Mplus approach. The results highlight the divergence of these techniques when sample sizes are small and/or multivariate kurtosis high. Recommendations include using multiple estimation techniques and, when sample sizes are large, sampling the data and reestimating the models to test both the robustness of the specifications and to quantify, to some extent, the large sample bias inherent in the χ ² test statistic.     

Advanced supplier partnership practices: a case study

This article describes how a supplier partnership was set up to avoid the typical purchasing relationship--price being inversely proportional to quantity and having the purchaser take all the risk of product obsolescence. The case study also describes how rate-based replenishment replaced time-based delivery, and how all these advantages were achieved at reduced administrative costs.     

Supplier involvement through certification: a case study

The Clarkson Company is located in the Reno-Sparks area of Nevada. We are a manufacturer of slurry valves for mining, pulp and paper, power generation, and various other industries. Because of the diverse nature of the applications for which the valves are used, the product can be configured a vast number of ways. Therefore, we use the assemble-to-order methodology for our master production scheduling. It became apparent over a decade ago that the use of materiel requirements planning (MRP) would aid us in controlling our low volume, high option mix environment. That journey led to The Clarkson Company becoming a Class A user of MRPII in June of 1990.     

Decentralisation of collective bargaining: a case study

While a number of surveys have identified a trend towards decentralised bargaining, this article is one of the first to examine the process of decentralisation. The decision of Coats Viyella PLC to decentralise bargaining in its apparel division is examined in detail and the outcomes of the first round of decentralised negotiations are analysed.     

The reversal of large stock price declines: The case of large firms

This paper examines the long-run reversal pattern for a sample of large U.S. firms that experienced significant stock price declines of more than 20 percent during a specific month. The results from the analysis are largely consistent with the overreaction hypothesis and significantly greater in magnitude than those reported by previous studies. Six and 12 months after their initial price decline, the stocks of large firms earn approximately 4 and 12 percent in excess of what was expected, respectively. However, the magnitude and trend of that reversal differs substantially across industries. Technology stocks experience the largest and strongest reversal pattern followed by manufacturing stocks, while service industry stocks exhibit a clear downward drift that lasts up to three years and can be described as investorunderreaction to the large price drop.     

Estimates of the return to schooling in Sweden from a large sample of twins

A large sample of twins was used to examine whether conventional estimates of the return to schooling in Sweden are biased because ability is omitted from the earnings–schooling relationship. Ignoring measurement error, the results indicate that omitting ability from the earnings–schooling relationship leads to estimates that are positively biased. However, reasonable estimates of the measurement-error-adjusted returns are both above and below the unadjusted estimates, showing that the results depend crucially on a parameter not known at this time. However, an estimate of the reliability ratio was obtained using two measures on educational attainment. With this estimate of the reliability ratio, the measurement-error-adjusted estimate of the return to schooling in the sample of identical twins indicates that there is at most a slight ability bias in the conventional estimates of the return to schooling. The fundamental assumption of this kind of study is that within-pair differences in educational attainment are randomly determined. This assumption was also tested, but no strong evidence to reject it was found.     

The realities of empowering teams: a case study

This case describes the transition of a manufacturing unit from operating under a rigid hierarchical environment to being run by empowered process management teams. This article traces the evolution of the transformation from the early days of skepticism and nonparticipation to the present, where most of the teams function with a large degree of autonomy and have made significant improvements in quality, lead time, and cost. Elements that help the team process are described, and potential hazards are outlined.     

Teamwork: a case study on development and performance

The most popular model of team development in Dutch socio-technical literature is a linear approach, which states that teams develop in four successive phases. A method for defining the particular phase a team is in was developed a number of years ago and was recently used in a large-scale survey at Volvo's cab manufacturing plant in Umeå (northern Sweden). Thirty-seven semi-autonomous teams were studied at this plant during a seven-month period. This paper examines the development of the teams and addresses the effects of team development on overall team performance. The aspect of team development was correlated to both performance in terms of quality of working life (QWL) and business performance (BP), which is an empirically unexplored field within team literature. The linear phase approach of team development could not be proved. Nevertheless, teams were found to develop in four important areas, with each aspect significantly affecting team performance.     

List augmentation with model based multiple imputation: a case study using a mixed-outcome factor model

This study concerns list augmentation in direct marketing. List augmentation is a special case of missing data imputation. We review previous work on the mixed outcome factor model and apply it for the purpose of list augmentation. The model deals with both discrete and continuous variables and allows us to augment the data for all subjects in a company's transaction database with soft data collected in a survey among a sample of those subjects. We propose a bootstrap-based imputation approach, which is appealing to use in combination with the factor model, since it allows one to include estimation uncertainty in the imputation procedure in a simple, yet adequate manner. We provide an empirical case study of the performance of the approach to a transaction data base of a bank.