FIMIX-PLS Calculation of aggregated solution

m.stolper · Post by **m.stolper** » Wed Apr 18, 2007 8:37 am

I have a short question concerning the calculation of the aggregated solution (k=1) with FIMIX-PLS. There occurs an error "Segments must be >1". Isn´t it possible to calculate the aggregated solution?

cringle · Post by **cringle** » Wed Apr 18, 2007 12:15 pm

Hi,

it is possible - but not really useful from our point of view. That's why we restricted this option. You do not interpret the results as in FIMIX-SEM. FIMIX-PLS indicates whether or not heterogeneity might be a problem that may be treated by segmentation. Thus, you are looking for two or more latent classes. However, if you still need the one segment solution, it is not very complicated to obtain the outcomes but requires a little effort. For the one segment solution, FIMIX-PLS results are identical to the usual PLS estimates - computation of the lnL is straightforward with a calculator.

Best regards
Christian