Message/Author 


...is there some way of finding out what units at Level 2 end up being used in an analysis?.. .one of my students is doing a multilevel sem...she has 76 units in her data set but the program only used 61 am assuming b/c missing data... but we now need to do some other analyses in SPSS but need to do them using only the 61 units used in the model test so are comparing apples to apples... is there some sort of descriptive command that you can enter that spits out the codes for each unit used in the data analysis like you can do in SPSS? 


You can save the data and see which cluster id's are represented in the saved data and use that data with SPSS. I think, however, the more important issue is why all 76 units are not being used. Please send the input, data, output, and your license number to support@statmodel.com if you want help on this. 


thanks Linda..16 units had missing data on one of the control variables so were not used... this was a multilevel SEM done by a colleague who is a statistician familar with Mplus but wasn't sure re how to id the units used in the analysis in Mplus programming...he used Version 5 I think.. ..so I ended up manually deleting cases from the SPSS file with no data for that control variable...but was wondering if it could all be done in Mplus to make life simpler... I think my Mplus program is no longer useable since it is the 4.0 version bought last year I think...I have not learned how to run it yet but would like to you since it is so much more powerful than AMOS and LISREL... 


You could have analyzed the saved data which would contain only the clusters used in the analysis. Version 4 came out in February 2006. There have been many changes since then. 


thanks so much...we will try this... 

Back to top 