PROCESS is telling me, that When the moderator is dichotomous, this is a test of equality of the. My predictor and moderator are dichotomous. I'm using PROCESS to test my moderated mediation. In earlier releases of version 3, using the multicategorical. A multicategorical variable has three or more categories. You should NOT use the multicategorical option with a dichotomous independent variable or moderators. Because the mathematics is the same for dichotomous X, W, and/or Z as it is for continuous variables, just put your dichotomous variable(s) in the PROCESS command or menu for X, W, and/or Z. Th Answer: PROCESS has always allowed dichotomous independent variables and moderators. Consider the equations: Y = cX + E 1 M = aX + E 2 Y = bM + c'X + E 3 where X is the causal variable, Y the outcome, and M the mediator and each is a dichotomy. The interested reader should consult their paper for more details. This note is on the testing of mediation using logistic regression and is largely based on a paper by MacKinnon and Dwyer (1993).
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |