I have investigated with multilevel mediation analysis. and It seemed to be well-performed. but, There is some problem Sum of Direct effect and Indirect effect is not equal to Total Effect. (For exampe, the size of direct Effect is 0.7409. and Indirect is 0.043, but The Size of total effect is 0.7495).
So I thought that It is because of Binary(Or Dichotomous) Outcomes. At the result of searching, There are many discussion how statisticians analyse the mediation model with binary outcomes or mediator.
hence, I want to ask to you some question.
1. How can I analyse multilevel mediation with binary(or dichotomous) variables in SAS?
2. Why didn’t work Mediation model with binary variable?
3. How do you think about the reason why the sum of Direct effect and Indirect effect is not equal to Total Effect? is it because of Binary variables?
If you have the answer or another thoughts, then Please Reply Answer.
Thank you.
In my research, I did Binary logistic regression. This is used to predict. It has shown a new direction in my research to predict the respondents who may wait to shop grocery during the opening hours. "Willingness to wait" is binary data. Yes=1. No =0.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis