Options

# "Logistic regression"

Hi,

I am working on Logistic Regression. In other statistical tools the value of dependent variables are 0 and 1. So we are getting P(y=1/x=x). But in RM

the dependent variable should be nominal (dichotomy ie true or false type ) . I am getting the result in RM but I would like to know this probability is corresponding to 'true ' or ' false' ie P(y=true/x=x) or P(y=false/x=x).

Thanks,

Nancy

I am working on Logistic Regression. In other statistical tools the value of dependent variables are 0 and 1. So we are getting P(y=1/x=x). But in RM

the dependent variable should be nominal (dichotomy ie true or false type ) . I am getting the result in RM but I would like to know this probability is corresponding to 'true ' or ' false' ie P(y=true/x=x) or P(y=false/x=x).

Thanks,

Nancy

Tagged:

0

## Answers

2,531Unicornon of the values is internally mapped to 0, the other one to 1. Which value becomes which, depends on the order in the data set.

But why do you care? Since P(y=1/x=x) = 1 - P(y=0/x=x), where is the difference?

Greetings,

Sebastian