# Mixed effects logistic Regression

#### smh5154

##### New Member
I have what I believe is a mixed effects logistic regression. My outcome is injury (y/n) and I have subject level (age, gender, smoking) and Class level (\$ Funds, instructor experience, etc.). I am trying to preduct injury based on the subject AND class level data. My outcome is by subject though...each person gets a 1 if they were injured and 0 if they were not. Any suggestions on how to proceed. I think I've found information to support it being in the spectrum of mixed effects logistic regression but I don't really know how to proceed. I'd appreciate any help. Thank you!

#### victorxstc

##### Pirate
Yes it seems to be a case of binary mixed model logistic regression. Also you can use GEE as Wilbert suggested as well as a conditional binary logistic regression. For doing it in SPSS, you must have SPSS 19 or above and look at its tutorial for very detailed instructions to use "Mixed Models -> Generalized Linear Models"... The procedure you are dealing with can be a little bit tricky (even in all-GUI-based SPSS), so before that please let us know how much are you familiar with stats as itself and stats software. Also see this thread for a very helpful step by step tutorial of doing it in R plus several super hints on the dos and don'ts of the analysis...

#### smh5154

##### New Member
I am familiar with statistics i have an applied statistics degree and background and I use SPSS very frequently and SAS occasionally. I've just never modeled multi-level where the two levels have different variables. I am aware of the case of nesting with the typical example of testing student scores but each student is within 1 of 5 teachers. But this is different because I am interested in the higher level of class/group and the class group predictors are different from the student level predictors. I really just don't know where to start and really what to expect my output to look like. The link I attached in my first message would be perfect if they showed step by step for example 1, but they showed it for example 2.