Multivariate logistic regression?

#1
Hi! With my master thesis project, I only use categorical data. I have one independent variabele, with four categories. My dependent variables are dichotomous (yes/no). I would like to use logistic regression to calculate the odds that when the categories of my independent variable are available, the independent variable will be present or not. I already succeeded so far. However, my dependent variables are psychological symptoms, so I have over 40 of them. If I perform logistic regressions for every symptom (dependent variable) one by one, the error will be big. Therefore, I would like to use one statistical test to predict multiple dependent variables at the same time. I know this is possible using a MANOVA, however, for a MANOVA the dependent variable needs to be at the ratio/interval level, which is not the case. So, I am wondering whether there is a possibility to make my multiple regression multivariate as well?
 

Karabiner

TS Contributor
#2
What is the topic of your study, what are the research questions?
How were the data collected and how large is your sample size?
Why do you perform logistic regression for 4*2 tables where simple
Chi² seems sufficient?

With kind regards

Karabiner