P-value for t test in GLM

Asked Jan 20 '18 at 23:35

Active Jul 22 '23 at 22:53

Viewed 394 times

I understand how the p-values are calculated for the coefficients in a GLM with known dispersion parameter. I can find rough estimates for the p-values using the Z table. However, when the GLM has estimated parameter, we perform a t-test. It seems that the p-values can again be found in the Z table. Why is this? Why do we not use the t table?

Also, we use a critical value from the t table to see whether our variable is significant for a t-test. I understand this, but then I don't understand why we switch to a Z table for the p-value.

edited Jan 20 '18 at 23:41

asked Jan 20 '18 at 23:35

2

In 2018, no one should be using tables. – Matthew Drury Jan 20 '18 at 23:39
@MatthewDrury In the exam we are only given tables, hence this is why I am trying to learn it this way. – Jan 20 '18 at 23:40
1

Thats tragic. :( – Matthew Drury Jan 20 '18 at 23:41

P-value for t test in GLM

0 Answers0