Рет қаралды 4,419
FULL EXPLANATION OF INNER MODEL ANALYSIS ON SMART PLS 3.0
The Fit Model Test is used to assess the compatibility between the observed correlations, where the Standardized Root Mean Square (SRMR) value is less than 0.10 or 0.08, then the model is considered suitable (see Hu and Bentler, 1999).
Then the model is said to be fit if the RMS Theta or Root Mean Square Theta value is less than 0.102.
• Normal Fit Index (NFI) value between 0 and 1. the closer to 1 the better/the model is more suitable.
The assumption or requirement in the inner least square partial model analysis is that there is no multicollinearity problem, namely there is a strong intercorrelation between latent variables.
• Multicollinearity is a phenomenon in which two or more independent variables or exogenous constructs are highly correlated so that the predictive ability of the model is not good (Sekaran and Bougie, 2016).
Test conditions: The VIF value must be less than 5, because if it is more than 5 it indicates the existence of collinearity between constructs (Sarstedt et al., 2017).
The evaluation of the model can also be seen from the R-Square value, where this R Square is also a goodness fit test for the PLS inner SEM model.
We are going to Slide
• Inner model is used to predict the relationship between latent variables.
• Used to measure model quality criteria or goodness of fit
• The coefficient of determination gives an indication of the magnitude of the effect of the exogenous latent variable on the amount of the endogenous latent.
• The value of the coefficient of determination can be seen from the R Square value for endogenous latent constructs as predictive power.
• The value of the coefficient of determination (R Square) between 0 and 1, the value of R Square closer to 1 the model the better or feasible.
Meanwhile, Adjusted R Square is the corrected R Square value based on the standard error value. The value of Adjusted R Square provides a stronger picture than R Square in assessing the ability of an exogenous construct to explain endogenous constructs.
• Path Coefficients are used to determine the magnitude of the effect partially and indicate the direction of the variable relationship, whether the relationship between variables is positive or negative.
• Path Coefficients range in value from -1 to 1.
• GoF value to test the overall fit of the model, both for the outer model and inner model, whether there is a match with the observed value with the expected value in the model.
#InnerModel Explanation
#AnalisisInnerModel
#InterpretasiOutputSmartPLS