Hi there!
We are doing research work in the domain of business and management, and our research model is as follows:
X --> M1 --> M2 --> M3 --> Y, where
X, the IV, is continuous (scale) variable
M1 and M2 are categorical mediators
M3 is a continuous (scale) mediator
Y, the DV, is a continuous (scale) variable
Right now, we are at a loss as to how to test this model, i.e., which software to use. Unfortunately, Hayes's SPSS PROCESS Macro cannot accommodate the mediation of categorial mediators yet, so that's ruled out.
Do you think that MPlus could be a good way of testing this model?
Or, can you suggest an R package that can potentially test this?
In R, the "lavaan" package is commonly used for structural equation modeling, and it does support mediation analyses. The "mediation" package in R is another option that specifically focuses on mediation analysis.
Mplus can handle categorical dependent variables (mediators and outcomes) in path models. See
https://www.statmodel.com/Mediation.shtml
https://www.statmodel.com/download/causalmediation.pdf
Hi Hewa Ghafor Hassan , thank you for taking the time to answer my question. Specifically, we want to test a serial mediation model that has two categorical mediators, as I have discussed in the question text. I'll check whether Lavaan can accommodate such a model.
Testing for serial mediation with two categorical mediators involves examining the sequential mediation effects of two categorical variables on the relationship between an independent variable and a dependent variable. Here's a step-by-step guide on how you might conduct such an analysis using Mplus:
1. Define Variables:
VARIABLE:
NAMES = x1 x2 m1 m2 y;
CATEGORICAL = m1 m2 y;
Define your variables (x1, x2, m1, m2, and y), and specify that m1, m2, and y are categorical.
2. Specify the Model:
MODEL:
%OVERALL%
y ON x1 x2;
%m1%
m1 ON x1;
%m2%
m2 ON x2 m1;
%y%
y ON m2;
3. Specify Indirect Effects:
MODEL INDIRECT: m1 IND x1 IND m2; m2 IND m1 IND x2; y IND m2 IND m1 IND x1;
4. Run the Analysis:
ANALYSIS: TYPE = GENERAL; ESTIMATOR = WLSMV;
% or use other appropriate estimator
5. Interpret Results:
Examine the output for the indirect effects. If the confidence intervals for the indirect effects do not include zero, you can infer that there is evidence of significant mediation.
OUTPUT: TECH1 TECH8;
You may want to request specific output to obtain detailed information on the indirect effects.
Use SEM. You can use any of the software such as MPlus, AMOS, R-package Lavaan
Check this paper
Sarstedt, M., Hair Jr, J. F., Nitzl, C., Ringle, C. M., & Howard, M. C. (2020). Beyond a tandem analysis of SEM and PROCESS: Use of PLS-SEM for mediation analyses!. International Journal of Market Research, 62(3), 288-299.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size