|
|
Specification of ran. slope - weird e... |
|
Message/Author |
|
|
Dear Sir or Madam, after successfully running a two-level model with random intercept and predictor variables on both levels, I now want to run the same model but with random slopes and test the slope variability. I followed the instructions of the UserGuide but MPlus aborts and tells me "A latent variable declared on the between level cannot be used on the within level - Problem with variables SLOPE_1 and SLOPE_2". But these variables are the random slopes of regressions on the within level and the predictors variables of these regressions ARE specified as variables on the within level! The dependent variable is measured on within level but not defined as within variable as it is supposed to have variance on both levels. I do not understand what I did wrong and would be very grateful for your help! Here the relevant part of my syntax: usevariables = ZD_C NA_C ERHOLT_T ZD_T FATIG_T; ! FATIG_T is the dependent variable and was measured on L1 cluster = VPN_NR; within = ERHOLT_T ZD_T; between = ZD_C NA_C; define: center ZD_C NA_C (grandmean); center ERHOLT_T ZD_T (groupmean); analysis: type = twolevel random; estimator = ml; model: %within% SLOPE_1 | FATIG_T on ZD_T; SLOPE_2 | FATIG_T on ERHOLT_T; FATIG_T with SLOPE_1; FATIG_T with SLOPE_2; %between% FATIG_T on NA_C ZD_C; |
|
|
Your error is FATIG_T with SLOPE_1; FATIG_T with SLOPE_2; You cannot estimate within-level parameters related to the random slopes because they vary only on Between. |
|
|
Dear Prof. Muthén, thank you very much for your fast and helpful response! Without these commands MPlus had no problems running my analysis :-). I beg your pardon that I made such a stupid mistake! Have a nice day, Kerstin Mertz |
|
Back to top |
|
|