@phdthesis{Schuberth2019,
  author    = {Schuberth, Florian},
  title     = {Composite-based Methods in Structural Equation Modeling},
  doi       = {10.25972/OPUS-15465},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-154653},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2019},
  abstract  = {This dissertation deals with composite-based methods for structural equation models with latent variables and their enhancement. It comprises five chapters. Besides a brief introduction in the first chapter, the remaining chapters consisting of four essays cover the results of my PhD studies.Two of the essays have already been published in an international journal. The first essay considers an alternative way of construct modeling in structural equation modeling.While in social and behavioral sciences theoretical constructs are typically modeled as common factors, in other sciences the common factor model is an inadequate way construct modeling due to its assumptions. This essay introduces the confirmatory composite analysis (CCA) analogous to confirmatory factor analysis (CFA). In contrast to CFA, CCA models theoretical constructs as composites instead of common factors. Besides the theoretical presentation of CCA and its assumptions, a Monte Carlo simulation is conducted which demonstrates that misspecifications of the composite model can be detected by the introduced test for overall model fit. The second essay rises the question of how parameter differences can be assessed in the framework of partial least squares path modeling. Since the standard errors of the estimated parameters have no analytical closed-form, the t- and F-test known from regression analysis cannot be directly used to test for parameter differences. However, bootstrapping provides a solution to this problem. It can be employed to construct confidence intervals for the estimated parameter differences, which can be used for making inferences about the parameter difference in the population. To guide practitioners, guidelines were developed and demonstrated by means of empirical examples. The third essay answers the question of how ordinal categorical indicators can be dealt with in partial least squares path modeling. A new consistent estimator is developed which combines the polychoric correlation and partial least squares path modeling to appropriately deal with the qualitative character of ordinal categorical indicators. The new estimator named ordinal consistent partial least squares combines consistent partial least squares with ordinal partial least squares. Besides its derivation, a Monte Carlo simulation is conducted which shows that the new estimator performs well in finite samples. Moreover, for illustration, an empirical example is estimated by ordinal consistent partial least squares. The last essay introduces a new consistent estimator for polynomial factor models. Similarly to consistent partial least squares, weights are determined to build stand-ins for the latent variables, however a non-iterative approach is used. A Monte Carlo simulation shows that the new estimator behaves well in finite samples.},
  subject      = {Strukturgleichungsmodell},
  language  = {en}
}
@phdthesis{Rademaker2020,
  author    = {Rademaker, Manuel Elias},
  title     = {Composite-based Structural Equation Modeling},
  doi       = {10.25972/OPUS-21593},
  url       = {http://nbn-resolving.de/urn:nbn:de:bvb:20-opus-215935},
  school      = {Universit{\"a}t W{\"u}rzburg},
  year      = {2020},
  abstract  = {Structural equation modeling (SEM) has been used and developed for decades across various domains and research fields such as, among others, psychology, sociology, and business research. Although no unique definition exists, SEM is best understood as the entirety of a set of related theories, mathematical models, methods, algorithms, and terminologies related to analyzing the relationships between theoretical entities -- so-called concepts --, their statistical representations -- referred to as constructs --, and observables -- usually called indicators, items or manifest variables. This thesis is concerned with aspects of a particular strain of research within SEM -- namely, composite-based SEM. Composite-based SEM is defined as SEM involving linear compounds, i.e., linear combinations of observables when estimating parameters of interest. The content of the thesis is based on a working paper (Chapter 2), a published refereed journal article (Chapter 3), a working paper that is, at the time of submission of this thesis, under review for publication (Chapter 4), and a steadily growing documentation that I am writing for the R package cSEM (Chapter 5). The cSEM package -- written by myself and my former colleague at the University of Wuerzburg, Florian Schuberth -- provides functions to estimate, analyze, assess, and test nonlinear, hierarchical and multigroup structural equation models using composite-based approaches and procedures. In Chapter 1, I briefly discuss some of the key SEM terminology. Chapter 2 is based on a working paper to be submitted to the Journal of Business Research titled "Assessing overall model fit of composite models in structural equation modeling". The article is concerned with the topic of overall model fit assessment of the composite model. Three main contributions to the literature are made. First, we discuss the concept of model fit in SEM in general and composite-based SEM in particular. Second, we review common fit indices and explain if and how they can be applied to assess composite models. Third, we show that, if used for overall model fit assessment, the root mean square outer residual covariance (RMS_theta) is identical to another well-known index called the standardized root mean square residual (SRMR). Chapter 3 is based on a journal article published in Internet Research called "Measurement error correlation within blocks of indicators in consistent partial least squares: Issues and remedies". The article enhances consistent partial least squares (PLSc) to yield consistent parameter estimates for population models whose indicator blocks contain a subset of correlated measurement errors. This is achieved by modifying the correction for attenuation as originally applied by PLSc to include a priori assumptions on the structure of the measurement error correlations within blocks of indicators. To assess the efficacy of the modification, a Monte Carlo simulation is conducted. The paper is joint work with Florian Schuberth and Theo Dijkstra. Chapter 4 is based on a journal article under review for publication in Industrial Management \& Data Systems called "Estimating and testing second-order constructs using PLS-PM: the case of composites of composites". The purpose of this article is threefold: (i) evaluate and compare common approaches to estimate models containing second-order constructs modeled as composites of composites, (ii) provide and statistically assess a two-step testing procedure to test the overall model fit of such models, and (iii) formulate recommendation for practitioners based on our findings. Moreover, a Monte Carlo simulation to compare the approaches in terms of Fisher consistency, estimated bias, and RMSE is conducted. The paper is joint work with Florian Schuberth and J{\"o}rg Henseler.},
  subject      = {trukturgleichungsmodell},
  language  = {en}
}