Message/Author 

dlsp posted on Friday, April 23, 2010  6:25 am



Dear Dr. Muthèn, I am running a model X > y > u > z. x, y, and z are continuous variables. u is a binary variable with a distribution of 15/85. Is there a possibility to correct for a basrate of 50/50 in Mplus? 


Perhaps you are thinking of some sort of poststratification adjustment using weights. I don't know how that would be done in this case. 

dlsp posted on Friday, April 23, 2010  10:21 am



Is there a possibility to correct the correlation matrix by hand or to impute a corrected correlation matrix? If I use the weighting option would you use the weight or frequency weigth option? Would you weight the low probability cases in the 15/85 distribution with a weighting factor of 5.66 and the other cases with a weighting factor of 1? 


I don't know enough about this to say, I'm afraid. 

Back to top 