China
Africa
Explore chapters and articles related to this topic
The role of social media marketing and environmental knowledge on green skincare purchase intention
Published in Siska Noviaristanti, Contemporary Research on Management and Business, 2023
The present study used a non-probability sampling technique using a purposive method. Data were collected by an online questionnaire to 372 respondents in Indonesia and 356 usable questionnaires were obtained after removal of the outliers. Only social media active users and consumers who had never bought green skincare were allowed to complete the questionnaire. All measurement items on questionnaires were obtained from previous research and measured using a five-point Likert scale (1=Strongly Disagree, 5=Strongly Agree). The collected data were analyzed using structural equation modeling (SEM) with Amos 26.0 software package. Firstly, the outliers were removed to obtain clean data before further analysis and a normality test was performed using SPSS 25 to achieve a normal distribution result for SEM analysis. In SEM analysis, the researchers conducted measurement model testing to perform CFA and structural model testing to analyze the relationship between variables. In this study, the researchers examined the direct and indirect effect of variables to better understand which variables showed a significant effect toward green skincare products purchase intention.
Models and Modelling
Published in Shyama Prasad Mukherjee, A Guide to Research Methodology, 2019
Models to represent structural relations connecting elements and their features within a system were developed through interactions among social scientists (including economists), geneticists and statisticians. And many problems in choosing and fitting these models and even in interpreting the results of analysis based on such models attract statisticians. Structural equation modelling (SEM) is a statistical exercise for testing and estimating causal relations using a combination of statistical data (observed as well as unobservable) and qualitative causal assumptions. SEM allows both exploratory and confirmatory analysis, suiting theory development and also theory testing and validation. Path analysis, factor analysis and linear regression analysis are all special cases of SEM.
Overview on Structural Equation Modeling
Published in Sergey V. Samoilenko, Kweku-Muata Osei-Bryson, Creating Theoretical Research Frameworks Using Multiple Methods, 2017
Sergey V. Samoilenko, Kweku-Muata Osei-Bryson
The purpose of this chapter is to offer a general, necessarily brief and nontechnical, overview of structural equation modeling (SEM)—a methodology representing the second generation of multivariate analysis. Unlike the statistical tools of the first generation, exemplified by such techniques as cluster analysis, multivariate regression, principal component analysis (PCA), and others, SEM allows for answering multiple interrelated research questions within a single analysis. Use of SEM allows researchers to posit the presence of the relationships between the multiple unobserved, or latent, variables, where every latent variable is associated with multiple observed variables, often called indicators or measures. The corresponding basic structure of SEM is illustrated by Figure 8.1.
Social media use, loneliness and psychological distress in emerging adults
Published in Behaviour & Information Technology, 2023
Zoe Taylor, Ala Yankouskaya, Constantina Panourgia
We estimated the initial model’s fit where residuals associated with multiple predictors and outcomes were permitted to covary. The model showed reasonably good model fit according to multiple SEM fit statistics and indices: Root Mean Square Error of Approximation (RMSEA) = .08, 95% CI [0.001, 0.10]; Comparative fit index (CFI) = .982; Tucker-Lewis index (TLI) = .95 (rule of thumb guidelines are that CFI ≥ .95, TLI ≥ .95 represent a good fitting model). It has to be noted that although previous research proposed a stringent cut-off value for RMSEA of 0.06 (Hu and Bentler 1999) or the upper limit of less than 0.08 (McQuitty 2004), recent studies argued for flexible cut-offs (Niemand and Mai 2018). This is particularly relevant to SEM, which considers only the theoretically relevant paths. Therefore, the model fit metrics suggest that our theoretically motivated model of the covariance among variables provided a reasonably good approximation of the data obtained in this study (additional fit metrics are presented in Supplementary Material, Note 3).
The influence of distractions of the home-work environment on mental health during the COVID-19 pandemic
Published in Ergonomics, 2023
Lisanne Bergefurt, Rianne Appel-Meulenbroek, Celine Maris, Theo Arentze, Minou Weijs-Perrée, Yvonne de Kort
First, bivariate analyses were conducted to get insights into the significance of direct relationships between variables (Field 2013). Both internal, as well as external relationships, were tested, except for the internal relationships between personal characteristics. The significant internal and external relationships were then used as input for the path analysis. Path analysis is a special case of structural equation modelling (SEM), in which multiple direct and indirect relationships between independent and dependent variables are determined simultaneously. All relationships that were found to be insignificant at the 0.05 (t < 1.96) significance level were deleted from the path model to overcome the risk of an overfitted model. This backward stepwise process was repeated until an acceptable model was found and all insignificant relationships were removed from the path model (Streiner 2005; Hu, Bentler, and Hu 1999). The statistical package Lavaan was used in RStudio to conduct the path analysis (Rosseel 2012).
Adoption patterns of autonomous technologies in Logistics: evidence for Niagara Region
Published in Transportation Letters, 2022
Amir Mohammadian Amiri, Mark R. Ferguson, Saiedeh Razavi
Path analysis, as another component of SEM, seeks to investigate causal models by exploring the relationships between a dependent variable and two or more independent variables. This technique can be considered as a specific form of SEM that contains only observed variables. Compared to SEM, path analysis has more restrictive assumptions. Path analysis measures all variables without considering error, while SEM uses latent variables to account for measurement error (Xue 2006). In other words, the first part of SEM, a structural model, relates to the relationship between endogenous and latent exogenous variables, which allows the assessment of both strength and direction of the causal effects among these variables, and the second component, measurement models, the relationship between the latent variables and their manifest or observable indicators (Golob 2003).