|
|
Cross Classified Multinomial Logit |
|
Message/Author |
|
Chris Weber posted on Sunday, August 26, 2012 - 8:06 am
|
|
|
Hello, I am attempting to fit a multinomial logit (3 categories), with random effects. The intercepts vary across news organizations and congressional races. However, news organizations are not perfectly nested within races, leading a cross classified structure. My question is whether this type of model can be fit in v6.12. If so, is the model I include below accomplishing this goal, or is it assuming observations are nested within news organization nested within races? I have not been able to find documentation on cross classified models in mplus Thanks, Chris USEVARIABLES ARE Margin days_elec race_id Org_Num offyear corporate female_candidate gov_race open_seat logcirc incumbent_years char femcanXgov; WITHIN= corporate days_elec logcirc Margin offyear female_candidate gov_race open_seat incumbent_years femcanXgov ; CLUSTER are race_id Org_Num; nominal is char; missing are all (999); analysis: type=complex twolevel; model: %within% char#1 char#2 on Margin days_elec corporate logcirc offyear female_candidate gov_race open_seat incumbent_years femcanXgov; %between% char#1 char#2 ; |
|
|
Cross-classified models are not available in Version 6.12. They will be available in Version 7. |
|
Back to top |
|
|