Confusion in determining significance of categorical variable using GLM function

categorical
logistic_regression

#1

Hi,

When trying to predict Dependents variable using Married categorical variable(with values YES and NO), passing through GLM function gives following output

Coefficients:
Estimate Std. Error z value Pr(>|z|)
(Intercept) -1.4933 0.1405 -10.62 <2e-16 ***
MarriedYes 1.7536 0.1623 10.81 <2e-16 ***

It doesn’t give any output for No value. What is the problem here ?

Thanks,
Vivek


#2

Hi @newton2304

The p-value for the variable is quite less. This means that the ‘variable’ Married is significant. Different levels won’t have different significance. Significance of a variable is meaningful in its own.
The matrix that you have got is correct.

See this:
http://www.r-tutor.com/elementary-statistics/logistic-regression/significance-test-logistic-regression


#3

So although its showing pvalue with YES value , the variable itself is significant. Right ?


#4

Yes, thats right.


#5

Hi @newton2304

Well when married NO the result will be the intercept as you will have Y= Intercept + 0 * Married Yes

Hope this help

Alain