scholarly journals Prior Sensitivity in Bayesian Structural Equation Modeling for Sparse Factor Loading Structures

2020 ◽  
Vol 80 (6) ◽  
pp. 1025-1058 ◽  
Author(s):  
Xinya Liang
Keyword(s):  
Structural Equation Modeling ◽  
Variable Selection ◽  
Structural Equation ◽  
Factor Loading ◽  
Model Fit ◽  
Equation Modeling ◽  
Model Parameters ◽  
Shrinkage Priors ◽  
Bayesian Structural Equation Modeling ◽  
The Impact

Bayesian structural equation modeling (BSEM) is a flexible tool for the exploration and estimation of sparse factor loading structures; that is, most cross-loading entries are zero and only a few important cross-loadings are nonzero. The current investigation was focused on the BSEM with small-variance normal distribution priors (BSEM-N) for both variable selection and model estimation. The prior sensitivity in BSEM-N was explored in factor analysis models with sparse loading structures through a simulation study (Study 1) and an empirical example (Study 2). Study 1 examined the prior sensitivity in BSEM-N based on the model fit, population model recovery, true and false positive rates, and parameter estimation. Seven shrinkage priors on cross-loadings and five noninformative/vague priors on other model parameters were examined. Study 2 provided a real data example to illustrate the impact of various priors on model fit and parameter selection and estimation. Results indicated that when the 95% credible intervals of shrinkage priors barely covered the population cross-loading values, it resulted in the best balance between true and false positives. If the goal is to perform variable selection, a sparse cross-loading structure is required, preferably with a minimal number of nontrivial cross-loadings and relatively high primary loading values. To improve parameter estimates, a relatively large prior variance is preferred. When cross-loadings are relatively large, BSEM-N with zero-mean priors is not recommended for the estimation of cross-loadings and factor correlations.

scholarly journals Prior sensitivity analysis in default Bayesian structural equation modeling

2017 ◽  
Author(s):  
Sara van Erp ◽  
Joris Mulder ◽  
Daniel L. Oberski
Keyword(s):  
Sensitivity Analysis ◽  
Structural Equation Modeling ◽  
Structural Equation ◽  
Empirical Bayes ◽  
Mean Squared Error ◽  
Small Samples ◽  
Equation Modeling ◽  
Model Parameters ◽  
Ml Estimation ◽  
Bayesian Structural Equation Modeling

Bayesian structural equation modeling (BSEM) has recently gained popularity because it enables researchers to fit complex models while solving some of the issues often encountered in classical maximum likelihood (ML) estimation, such as nonconvergence and inadmissible solutions. An important component of any Bayesian analysis is the prior distribution of the unknown model parameters. Often, researchers rely on default priors, which are constructed in an automatic fashion without requiring substantive prior information. However, the prior can have a serious influence on the estimation of the model parameters, which affects the mean squared error (MSE), bias, coverage rates, and quantiles of the estimates.In this paper, we investigate the performance of three different default priors: noninformative improper priors, vague proper priors, and empirical Bayes priors, with the latter being novel in the BSEM literature. Based on a simulation study, we find that these three default BSEM methods may perform very differently, especially with small samples. A careful prior sensitivity analysis is therefore needed when performing a default BSEM analysis. For this purpose, we provide a practical step-by-step guide for practitioners to conducting a prior sensitivity analysis in default BSEM. Our recommendations are illustrated using a well-known case study from the structural equation modeling literature and all code for conducting the prior sensitivity analysis is made available in the online supplemental material.

Factor Structure of the PANAS With Bayesian Structural Equation Modeling in a Chinese Sample

2021 ◽  
pp. 016327872199679
Author(s):  
Minsun Kim ◽  
Ze Wang
Keyword(s):  
High School Students ◽  
Structural Equation Modeling ◽  
Factor Structure ◽  
Structural Equation ◽  
Model Fit ◽  
Self Report ◽  
Equation Modeling ◽  
School Students ◽  
Chinese Sample ◽  
Bayesian Structural Equation Modeling

The Positive and Negative Affect Schedule (PANAS) is the most widely used self-report instrument for assessing affect. However, there are inconsistent findings regarding the factor structure of the PANAS. In this study, we applied Bayesian structural equation modeling (BSEM) to investigate the structure of the PANAS using data from a sample of 893 Chinese middle and high school students. Four models, the orthogonal two-, the oblique two-, the three-, and the bi-factor models were tested with prior specifications including approximately zero cross-loadings and residual covariances. The results indicated that the orthogonal two-factor model specified with informative priors for both cross-loadings and residual correlations has the best model fit. Confirmatory factor analysis with the maximum likelihood estimator (ML-CFA) based on modifications from BSEM analysis showed improved model fit compared to ML-CFA based on frequentist analysis, which is the evidence for the merit of BSEM for addressing misspecifications.

scholarly journals The Impact of Building Façade Reflectivity on Pedestrian Visual Comfort with the Application of Bayesian Structural Equation Modeling

2021 ◽  
Vol 25 (1) ◽  
pp. 211-219
Author(s):  
Norishahaini Mohamed Ishak ◽  
Hashem Salarzadeh Jenatabadi ◽  
Siti Nurul Ainun Mohd. Mustafa ◽  
Jamalunlaili Abdullah
Keyword(s):  
Structural Equation Modeling ◽  
Structural Equation ◽  
Equation Modeling ◽  
Visual Comfort ◽  
Building Facade ◽  
Bayesian Structural Equation Modeling ◽  
The Impact

Further Insights Into the German Version of the Multidimensional Assessment of Interoceptive Awareness (MAIA)

2019 ◽  
Vol 35 (3) ◽  
pp. 317-325 ◽  
Author(s):  
Dorota Reis
Keyword(s):  
Structural Equation Modeling ◽  
Structural Equation ◽  
Equation Modeling ◽  
Interoceptive Awareness ◽  
Multidimensional Assessment ◽  
New Instrument ◽  
Confirmatory Factor ◽  
Complex Models ◽  
Approximate Zero ◽  
Bayesian Structural Equation Modeling

Abstract. Interoception is defined as an iterative process that refers to receiving, accessing, appraising, and responding to body sensations. Recently, following an extensive process of development, Mehling and colleagues (2012) proposed a new instrument, the Multidimensional Assessment of Interoceptive Awareness (MAIA), which captures these different aspects of interoception with eight subscales. The aim of this study was to reexamine the dimensionality of the MAIA by applying maximum likelihood confirmatory factor analysis (ML-CFA), exploratory structural equation modeling (ESEM), and Bayesian structural equation modeling (BSEM). ML-CFA, ESEM, and BSEM were examined in a sample of 320 German adults. ML-CFA showed a poor fit to the data. ESEM yielded a better fit and contained numerous significant cross-loadings, of which one was substantial (≥ .30). The BSEM model with approximate zero informative priors yielded an excellent fit and confirmed the substantial cross-loading found in ESEM. The study demonstrates that ESEM and BSEM are flexible techniques that can be used to improve our understanding of multidimensional constructs. In addition, BSEM can be seen as less exploratory than ESEM and it might also be used to overcome potential limitations of ESEM with regard to more complex models relative to the sample size.

scholarly journals Structural equation modeling to shed light on the controversial role of climate on the spread of SARS-CoV-2

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Alessia Spada ◽  
Francesco Antonio Tucci ◽  
Aldo Ummarino ◽  
Paolo Pio Ciavarella ◽  
Nicholas Calà ◽  
...  
Keyword(s):  
Structural Equation Modeling ◽  
Population Density ◽  
Structural Equation ◽  
Climatic Factors ◽  
Demographic Data ◽  
Factor Loading ◽  
Statistical Technique ◽  
Global Scale ◽  
Equation Modeling ◽  
Relevant Role

AbstractClimate seems to influence the spread of SARS-CoV-2, but the findings of the studies performed so far are conflicting. To overcome these issues, we performed a global scale study considering 134,871 virologic-climatic-demographic data (209 countries, first 16 weeks of the pandemic). To analyze the relation among COVID-19, population density, and climate, a theoretical path diagram was hypothesized and tested using structural equation modeling (SEM), a powerful statistical technique for the evaluation of causal assumptions. The results of the analysis showed that both climate and population density significantly influence the spread of COVID-19 (p < 0.001 and p < 0.01, respectively). Overall, climate outweighs population density (path coefficients: climate vs. incidence = 0.18, climate vs. prevalence = 0.11, population density vs. incidence = 0.04, population density vs. prevalence = 0.05). Among the climatic factors, irradiation plays the most relevant role, with a factor-loading of − 0.77, followed by temperature (− 0.56), humidity (0.52), precipitation (0.44), and pressure (0.073); for all p < 0.001. In conclusion, this study demonstrates that climatic factors significantly influence the spread of SARS-CoV-2. However, demographic factors, together with other determinants, can affect the transmission, and their influence may overcome the protective effect of climate, where favourable.

Should I suggest this YouTube clip? The impact of UGC source credibility on eWOM and purchase intention

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Mazzini Muda ◽  
Muhammad Iskandar Hamzah
Keyword(s):  
Social Media ◽  
Structural Equation Modeling ◽  
Purchase Intention ◽  
Structural Equation ◽  
Word Of Mouth ◽  
Source Credibility ◽  
Equation Modeling ◽  
Content Type ◽  
The Social ◽  
The Impact

PurposeIn spite of the increasing organic and interactive marketing activities over social media, a general understanding of the source credibility of voluntary user-generated content (UGC) is still limited. In line with the social identity theory, this paper examines the effects of consumers' perceived source credibility of UGC in YouTube videos on their attitudes and behavioral intentions. Additionally, source homophily theory is included to predict the antecedent of source credibility.Design/methodology/approachThree hundred and seventy two Generation Y respondents were interviewed using snowball sampling. Data were analyzed with component-based structural equation modeling technique of partial least squares-structural equation modeling (PLS-SEM).FindingsFindings confirmed that perceived source credibility indirectly affects purchase intention (PI) and electronic word-of-mouth via attitude toward UGC. Besides, perceived source credibility mediates the effect of perceived source homophily on attitude toward UGC.Practical implicationsSince today's consumers have begun to trust and rely more on UGC than company-generated content on social media when making purchase decisions, companies may reconsider democratizing certain aspects of their branding strategies. Firms may fine-tune their marketing communication budgets – not only just by sponsoring public figures and celebrities but also by nurturing coproductive engagements with independent content creators who are ordinary consumers. Endowed with their imposing credibility, these micro-influencers and prosumers have high potentials to be uplifted to brand ambassadors.Originality/valueWhile consumers' purchase outcome can be measured easily using metrics and analytics, the roles of source homophily in stages leading up to the purchase is still elusive. Drawing on the rich theoretical basis of source homophily may help researchers to understand not only how credibility and attitude are related to PI but also how this nexus generates positive word of mouth among UGC followers within the social media circles.

scholarly journals Perilaku Pencarian Variasi: Ditinjau Dari Perspektif Psikologi Dan Imbasnya Pada Loyalitas Kesikapan Konsumen

2008 ◽  
Vol 1 (1) ◽  
Author(s):  
Iin Mayasari
Keyword(s):  
Structural Equation Modeling ◽  
Structural Equation ◽  
Equation Modeling ◽  
Hypotheses Testing ◽  
Variety Seeking ◽  
Attitudinal Loyalty ◽  
Optimum Stimulation Level ◽  
Stimulation Level ◽  
Psychology Perspective ◽  
The Impact

This study examines the model that explains the internal aspect as the stimulusi in influencing consumers to do variety seeking. The conceptual model is discussed by applying the psychology perspective of the optimum stimulation level and the impact on attitudinal loyalty. The number of questionnaires is 1100 exemplars and distributed to seven universities in Yogyakarta. However, the appropriate questionnaires to be further analyzed are 654 exemplars. The hypotheses testing uses the structural equation modeling.

Sign in / Sign up

Export Citation Format

Share Document