You make it look easy! Thanks.

Thank you a lot for this wonderful video. I have two questions: In my thesis, the scale that I adopted is mainly from the literature, with a few items from my qualitative research. My question is: should I do an EFA or CFA? According to your video, since I already have items related to the latent variable, I need to perform only a CFA. Am I right? Also, if I opt for CFA, which option should I follow please: Option 1:performing only CB-SEM, to assess the measurement model (calculate - CB SEM algorithm), and the structural model (calculate - CB SEM Bootstrapping)using SmartPLS . Option 2: Do a CFA with SPSS by specifying the number of factors in data reduction (not AMOS), and then use PLS-SEM to assess the measurement model again (calculate - PLS SEM algorithm)and the structural model (calculate - PLS SEM Bootstrapping). Thank you.

James, please I sincerely appreciate your efforts. I have a question, please. The literature is divided on the interaction of three variables, such as X, Y, and Z. The various perspectives are as follows: Some contend that the X-Z link is mediated by Y. According to others, X mediates the Y-Z path. Some contend that Y moderates the path X-Z. and some contend that X moderates the path Y-Z. Can I test each of these models on its own?

@Gaskination

23 күн бұрын

Sorry for the slow reply. Holiday weekend in the US :) If you mean three-way interaction, it is simply the dual moderation of two moderating variables on the path from X to Y. So, you have a main hypothesized effect (X-Y), then that path is moderated by both moderators.

I have a model with likert scale and categorical scale. The categorical scale as dependent varible. I'm confusing how to calculate the data. Could you please make the tutorial for this case 😢🙏

@Gaskination

22 күн бұрын

If your dependent variable is categorical with more than two categories, then you must create dummy variables for the outcome (or use a different method, like logistic regression or ANOVA). If your DV is binary, then you can include it as a single variable and just recognize that positive coefficients imply that positive values of the IV are more associated with the group coded with the higher values (e.g., 1 vs zero) and negative values are more associated with the lower coded group.