# Thread: p-value for logistic regression

1. ## p-value for logistic regression

How do I find the p-value for logistic regression. For example , for a glm output I :

z -value : 1.32 P ( > | Z | ) = 0.186696

If it were a normal regression would do : 2 ( 1 - P ( < | Z | ) = 2 (1-0.4066) = ???

The t-value = Estimate/ Std.Error...but P(>|t|) i don't understand.

2. ## Re: p-value for logistic regression

The p-value is .186696. That is already giving you the p-value for the two sided test.

If you did what you're suggesting for the output for the normal regression you're going to get something that is 1) wrong and 2) nonsensical (you can't have a p-value greater than 1).

3. ## The Following User Says Thank You to Dason For This Useful Post:

noetsi (06-27-2016)

4. ## Re: p-value for logistic regression

I understand it's STUPID BUT I WANT TO UNDERSTAND how to calculate it.

In some exercise it's proposed the output in this way:

Estimate Std.Error z value Pr(>|Z|)

X 0.28143 0.21314 1.320 ???

5. ## Re: p-value for logistic regression

Why would you want to understand how to calculate something that is wrong?

 Tweet