Nonlinear Structural Equation Models Research Articles

Development and psychometric testing of a new instrument to measure the caring behaviour of nurses in Italian acute care settings

To develop and psychometrically test the Italian-language Nurse Caring Behaviours Scale, a short measure of nurse caring behaviour as perceived by inpatients. Patient perceptions of nurses' caring behaviours are a predictor of care quality. Caring behaviours are culture-specific, but no measure of patient perceptions has previously been developed in Italy. Moreover, existing tools show unclear psychometric properties, are burdensome for respondents, or are not widely applicable. Instrument development and psychometric testing. Item generation included identifying and adapting items from existing measures of caring behaviours as perceived by patients. A pool of 28 items was evaluated for face validity. Content validity indexes were calculated for the resulting 15-item scale; acceptability and clarity were pilot tested with 50 patients. To assess construct validity, a sample of 2,001 consecutive adult patients admitted to a hospital in 2014 completed the scale and was split into two groups. Reliability was evaluated using nonlinear structural equationmodelling coefficients. Measurement invariance was tested across subsamples. Item 15 loaded poorly in the exploratory factor analysis (n=983) and was excluded from the final solution, positing a single latent variable with 14 indicators. This model fitted the data moderately. The confirmatory factor analysis (n=1018) returned similar results. Internal consistency was excellent in both subsamples. Full scalar invariance was reached, and no significant latent mean differences were detected across subsamples. The new instrument shows reasonable psychometric properties and is a promising short and widely applicable measure of inpatient perceptions of nurse caring behaviours.

Nexus between Human Values and Overall Satisfaction with the Urban Environment. A Non-Lineal Partial Least Square Path Modeling. Argentine Cities Case

In this paper, we examine the relationship between Human Values and the Urban Environment, and how values are behaviorally instantiated in different environments, through Non-Linear Structural Equation Model based on Partial Least Square. Beginning with a unified framework of the study of values, we analyze its organization, antecedents, covariates and presumed consequences that emerge from the data. Schwartz Model is used because it has a coherent and flexible structure with the measure instrument that has been evolved. Ten categories of Human Values and two dimensions are used to evaluate the Urban Environment. We will prove that the Human Value Conservation, through the dimensions: Conformity, Security and Tradition depends on the perceptions of the Urban Environment, specifically on the dimension Consumption/Esthetic & Social Cohesion. We postulate tentatively that the Human Value Self-Enhancement, through the subdimensions Achievement and Power, it is a variable, which in short and long time, impacts on the Perception of the Urban Environment, specifically on the Perception of the System of Generation and Distribution of Habitat Resource.

Analysis of Generalized Nonlinear Structural Equation Models by Using Bayesian Approach with Application

In this paper, Bayesian analysis is used in nonlinear structural equation models with two population of data and the Gibbs sampling method is applied for estimation and model comparison. Hidden continuous normal distribution (censored normal distribution) is used to solve the problem of ordered categorical data in Bayesian multiple group SEMs and compared with the method that treats ordered categorical variables as a continuous normal distribution. Statistical inferences, which involve the estimation of parameters and their standard errors, and residuals analyses for testing the posited model are discussed. The proposed procedure is illustrated using real data with the results obtained from the WinBUGS program.

Model comparison of linear and nonlinear Bayesian structural equation models with dichotomous data

ABSTRACTIn this article, dichotomous variables are used to compare between linear and nonlinear Bayesian structural equation models. Gibbs sampling method is applied for estimation and model comparison. Statistical inferences that involve estimation of parameters and their standard deviations and residuals analysis for testing the selected model are discussed. Hidden continuous normal distribution (censored normal distribution) is used to solve the problem of dichotomous variables. The proposed procedure is illustrated by a simulation data obtained from R program. Analyses are done by using R2WinBUGS package in R-program.

Reexamining Nonlinear Structural Equation Modeling Procedures: The Effect of Parallel and Congeneric Measures

The current study examines the performance of the extended unconstrained approach (EXUC) and the latent moderated structural equation modeling procedure (LMS) in situations where quadratic and interaction terms are tested simultaneously and investigates their limitations with regard to the employment of parallel and congeneric measures, relatively low indicator reliabilities, and relatively large numbers of indicators. By means of a Monte Carlo study, we found LMS to be the best option for testing multiple nonlinear effects given sufficient sample size (n ≥ 500) and normally distributed exogenous variables. Its advantages became more prominent when indicator reliabilities were heterogeneous and small. The EXUC was a viable option for estimating the model when indicators were parallel and exhibited large indicator reliabilities. An empirical example of the results is provided, and the relevance of measurement model characteristics to assess nonlinear relationships is discussed.

A narrative review and meta-analysis of service interaction quality: new research directions and implications

Purpose – This paper aims to conceptualize and empirically test the determinants of service interaction quality (SIQ) as attitude, behavior and expertise of a service provider (SP). Further, the individual and simultaneous effects of SIQ and its dimensions on important marketing outcomes are tested. Design/methodology/approach – The narrative review of extant research helps formulate a conceptual model of SIQ, which is investigated using the univariate and multivariate meta-analysis. Findings – There are interdependencies between drivers of SIQ that underlines the need to conceptualize service interaction as a dyadic phenomenon; use contemporary multilevel models, dyadic models, non-linear structural equation modeling and process studies; and study new and diverse services contexts. Meta-analysis illustrates the relative importance of the three drivers of SIQ and, in turn, their impact on consumer satisfaction and loyalty. Research limitations/implications – The meta-analysis is based on existing research, which, unfortunately, has not examined critical services or exigency situations where SIQ is of paramount importance. Future research will be tasked with diversifying to several important domains where SIQ is a critical aspect of perceived service quality. Practical implications – This study emphasizes that, although the expertise of an SP is important, firms would be surprised to learn that the attitude and behavior of their employees are equally important antecedents. In fact, there is a delicate balance that needs to be found; otherwise, attitudinal factors can have an overall counterproductive effect on consumer satisfaction. Originality/value – This paper provides an empirical synthesis of SIQ and opens up interesting areas for further research.

Marginal integration for nonparametric causal inference

We consider the problem of inferring the total causal effect of a single continuous variable intervention on a (response) variable of interest. We propose a certain marginal integration regression technique for a very general class of potentially nonlinear structural equation models (SEMs) with known structure, or at least known superset of adjustment variables: we call the procedure S-mint regression. We easily derive that it achieves the convergence rate as for nonparametric regression: for example, single variable intervention effects can be estimated with convergence rate $n^{-2/5}$ assuming smoothness with twice differentiable functions. Our result can also be seen as a major robustness property with respect to model misspecification which goes much beyond the notion of double robustness. Furthermore, when the structure of the SEM is not known, we can estimate (the equivalence class of) the directed acyclic graph corresponding to the SEM, and then proceed by using S-mint based on these estimates. We empirically compare the S-mint regression method with more classical approaches and argue that the former is indeed more robust, more reliable and substantially simpler.

Dimensional structure of bodily panic attack symptoms and their specific connections to panic cognitions, anxiety sensitivity and claustrophobic fears.

Previous studies of the dimensional structure of panic attack symptoms have mostly identified a respiratory and a vestibular/mixed somatic dimension. Evidence for additional dimensions such as a cardiac dimension and the allocation of several of the panic attack symptom criteria is less consistent. Clarifying the dimensional structure of the panic attack symptoms should help to specify the relationship of potential risk factors like anxiety sensitivity and fear of suffocation to the experience of panic attacks and the development of panic disorder. In an outpatient multicentre study 350 panic patients with agoraphobia rated the intensity of each of the ten DSM-IV bodily symptoms during a typical panic attack. The factor structure of these data was investigated with nonlinear confirmatory factor analysis (CFA). The identified bodily symptom dimensions were related to panic cognitions, anxiety sensitivity and fear of suffocation by means of nonlinear structural equation modelling (SEM). CFA indicated a respiratory, a vestibular/mixed somatic and a cardiac dimension of the bodily symptom criteria. These three factors were differentially associated with specific panic cognitions, different anxiety sensitivity facets and suffocation fear. Taking into account the dimensional structure of panic attack symptoms may help to increase the specificity of the associations between the experience of panic attack symptoms and various panic related constructs.

On the Performance of Likelihood-Based Difference Tests in Nonlinear Structural Equation Models

This article investigates likelihood-based difference statistics for testing nonlinear effects in structural equation modeling using the latent moderated structural equations (LMS) approach. In addition to the standard difference statistic TD, 2 robust statistics have been developed in the literature to ensure valid results under the conditions of nonnormality or small sample sizes: the robust TDR and the “strictly positive” TDRP. These robust statistics have not been examined in combination with LMS yet. In 2 Monte Carlo studies we investigate the performance of these methods for testing quadratic or interaction effects subject to different sources of nonnormality, nonnormality due to the nonlinear terms, and nonnormality due to the distribution of the predictor variables. The results indicate that TD is preferable to both TDR and TDRP. Under the condition of strong nonlinear effects and nonnormal predictors, TDR often produced negative differences and TDRP showed no desirable power.

Modeling Nonlinear Structural Equation Models: A Comparison of the Two-Stage Generalized Additive Models and the Finite Mixture Structural Equation Model

Researchers have devoted some time and effort to developing methods for fitting nonlinear relationships among latent variables. In particular, most of these have focused on correctly modeling interactions between 2 exogenous latent variables, and quadratic relationships between exogenous and endogenous variables. All of these approaches require prespecification of the nonlinearity by the researcher, and are limited to fairly simple nonlinear relationships. Other work has been done using mixture structural equation models (SEMM) in an attempt to fit more complex nonlinear relationships. This study expands on this earlier work by introducing the 2-stage generalized additive model (2SGAM) approach for fitting regression splines in the context of structural equation models. The model is first described and then investigated through the use of simulated data, in which it was compared with the SEMM approach. Results demonstrate that the 2SGAM is an effective tool for fitting a variety of nonlinear relationships between latent variables, and can be easily and accurately extended to models including multiple latent variables. Implications of these results are discussed.

A general non-linear multilevel structural equation mixture model.

In the past 2 decades latent variable modeling has become a standard tool in the social sciences. In the same time period, traditional linear structural equation models have been extended to include non-linear interaction and quadratic effects (e.g., Klein and Moosbrugger, 2000), and multilevel modeling (Rabe-Hesketh et al., 2004). We present a general non-linear multilevel structural equation mixture model (GNM-SEMM) that combines recent semiparametric non-linear structural equation models (Kelava and Nagengast, 2012; Kelava et al., 2014) with multilevel structural equation mixture models (Muthén and Asparouhov, 2009) for clustered and non-normally distributed data. The proposed approach allows for semiparametric relationships at the within and at the between levels. We present examples from the educational science to illustrate different submodels from the general framework.

Evaluation of model fit in nonlinear multilevel structural equation modeling.

Evaluating model fit in nonlinear multilevel structural equation models (MSEM) presents a challenge as no adequate test statistic is available. Nevertheless, using a product indicator approach a likelihood ratio test for linear models is provided which may also be useful for nonlinear MSEM. The main problem with nonlinear models is that product variables are non-normally distributed. Although robust test statistics have been developed for linear SEM to ensure valid results under the condition of non-normality, they have not yet been investigated for nonlinear MSEM. In a Monte Carlo study, the performance of the robust likelihood ratio test was investigated for models with single-level latent interaction effects using the unconstrained product indicator approach. As overall model fit evaluation has a potential limitation in detecting the lack of fit at a single level even for linear models, level-specific model fit evaluation was also investigated using partially saturated models. Four population models were considered: a model with interaction effects at both levels, an interaction effect at the within-group level, an interaction effect at the between-group level, and a model with no interaction effects at both levels. For these models the number of groups, predictor correlation, and model misspecification was varied. The results indicate that the robust test statistic performed sufficiently well. Advantages of level-specific model fit evaluation for the detection of model misfit are demonstrated.

Linking formal controls to motor carrier performance: Curvilinear and interaction effects

We examine the relationships between formal management controls (FMCs)—activity controls and output controls—and motor carriers’ operational and market performance. Using nonlinear structural equation modeling, we identify curvilinear and interaction effects of FMCs on carrier performance. We contribute to theory by providing evidence that the relationships between FMCs and performance are more complex than previously theorized. Namely, excessive activity control will adversely affect market performance, especially when output control is low. However, when output control is high, increasing activity control improves market performance—up to a point. Accordingly, we provide managerial guidance on appropriate levels of FMC use.

Partner Selection in Brand Alliances: An Empirical Investigation of the Drivers of Brand Fit

We investigate whether partners in a brand alliance should be similar or dissimilar in brand image to foster favorable perceptions of brand fit. Using a Bayesian nonlinear structural equation model and evaluations of 1,200 brand alliances, we find that the conceptual coherence in brand personality profiles predicts attitudes towards a brand alliance. More specifically, we find that similarity in Sophistication and Ruggedness and moderate dissimilarity in Sincerity and Competence result in more favorable brand alliance evaluations. Overall, we find that similarity effects are more pronounced than dissimilarity effects. Implications for brand alliance strategies and marketing managers are discussed.

Consistent Partial Least Squares for Nonlinear Structural Equation Models

Partial Least Squares as applied to models with latent variables, measured indirectly by indicators, is well-known to be inconsistent. The linear compounds of indicators that PLS substitutes for the latent variables do not obey the equations that the latter satisfy. We propose simple, non-iterative corrections leading to consistent and asymptotically normal (CAN)-estimators for the loadings and for the correlations between the latent variables. Moreover, we show how to obtain CAN-estimators for the parameters of structural recursive systems of equations, containing linear and interaction terms, without the need to specify a particular joint distribution. If quadratic and higher order terms are included, the approach will produce CAN-estimators as well when predictor variables and error terms are jointly normal. We compare the adjusted PLS, denoted by PLSc, with Latent Moderated Structural Equations (LMS), using Monte Carlo studies and an empirical application.

Combining Formal Controls to Improve Firm Performance

Formal controls are an essential part of organizational design, and are theorized to positively influence organizational performance. Mello and Hunt ([2009] Transportation Journal, 48:20–39) were the first to suggest using formal controls to influence truck drivers' behavior. Extending the salesperson control literature, they documented an underresearched method of formal control that we term technology control, whereby firms use on‐board and communications technology to influence drivers. We build on their work by investigating the contingent effects of formal controls that influence the behavior of truck drivers and thereby influence the operational performance of firms in the U.S. motor carrier industry. Our work integrates the monitoring aspect of agency theory with the theory of operant conditioning and the theory of psychological reactance to develop a rationale for hypothesized contingent effects of formal controls on motor carriers' operational performance. We collected primary data pertaining to the controls used to influence truck drivers' behavior from a large sample of U.S. motor carrier firms. The results of our analyses using nonlinear structural equation modeling suggest a complex set of relationships between formal controls and operational performance. Our results shed light on scenarios where using technology to monitor driver behavior can result in positive and/or negative organizational outcomes.

A Bayesian semiparametric dynamic two-level structural equation model for analyzing non-normal longitudinal data

Analyses of non-normal data and longitudinal data to study changes in variables measured repeatedly over time have received considerable attention in social and psychological research. This paper proposes a dynamic two-level nonlinear structural equation model with covariates for analyzing multivariate longitudinal responses that are mixed continuous and ordered categorical variables. To cope with the non-normal continuous data, the corresponding residual errors at both first-level and second-level models are modeled through a Bayesian semiparametric modeling on the basis of a truncated and centered Dirichlet process with stick-breaking priors. The first-level model is defined for measures taken at each time point nested within individuals for investigating their characteristics that vary with time; while the second level is defined for individuals to assess their characteristics that are invariant with time. An algorithm based on the blocked Gibbs sampler is implemented for estimation of parameters. An efficient model comparison statistic, namely the Lν-measure, is also introduced. Results of a simulation study indicate that the performance of the Bayesian semiparametric estimation is satisfactory. The proposed methodologies are applied to a real longitudinal study concerning cocaine use.

RESPONSIVE AND PROACTIVE MARKET ORIENTATION AND INNOVATION SUCCESS UNDER MARKET AND TECHNOLOGICAL TURBULENCE

The study investigates how market and technological changes in an organization's business environment moderate the relationships between responsive and proactive market orientation, innovation success, and market success of the organization. The respondents in the study were senior managers of companies operating in a Central European country. The Internet survey resulted in 441 usable questionnaires. Data were analyzed using a non-linear structural equation models with MPLUS5. The results provide support for distinguishing between the two complementary forms of market orientation, proactive and responsive. While proactive market orientation is a determinant of both innovation and market success of the organization, the impact of responsive market orientation on the innovation and market success is positive and significant only in a rapidly changing market environment. Companies can improve their innovation success and in turn market success by improving their proactive market orientation, i.e. by investing resources in exploring customer needs, customer problems with existing products and latent customer needs. The study contributes to the literature by examining the entire chain of relationships between market orientation, innovation success and market success by adopting both a responsive and proactive market orientation. It is the first study that examines these relationships in the context of companies from a European country and with consideration of market turbulence/changes.

Bayesian analysis of multi-group nonlinear structural equation models with application to behavioral finance

Structural equation models (SEMs) have been widely used to determine the relationships among certain observed and latent variables in behavioral finance. The purpose of this paper is to develop a Bayesian approach for analysing multi-group nonlinear SEMs. Using recently developed tools in statistical computing, such as the Gibbs sampler, we propose an efficient method to estimate parameters and select an appropriate model. The proposed method is used to investigate the relationships among all identified influential factors that have an impact on the motivation for insider trading within the framework of behavioral finance.

Bayesian Lasso for Semiparametric Structural Equation Models

There has been great interest in developing nonlinear structural equation models and associated statistical inference procedures, including estimation and model selection methods. In this paper a general semiparametric structural equation model (SSEM) is developed in which the structural equation is composed of nonparametric functions of exogenous latent variables and fixed covariates on a set of latent endogenous variables. A basis representation is used to approximate these nonparametric functions in the structural equation and the Bayesian Lasso method coupled with a Markov Chain Monte Carlo (MCMC) algorithm is used for simultaneous estimation and model selection. The proposed method is illustrated using a simulation study and data from the Affective Dynamics and Individual Differences (ADID) study. Results demonstrate that our method can accurately estimate the unknown parameters and correctly identify the true underlying model.