Interpreting test values in GLM (Python Statsmodels)

f_ie

New Member
#1
Hi!
Being fairly new to regression models, I have a few questions regarding interpretation of a GLM summary presented below. I have 18 independent variables x1-x22 and two dependent variables: y1, y2. Do P>|z| being 0.000 for the presented variables x1-x4 imply that these variables are statistically significant in forecasting my dependent variables: y1 & y2 (Binomial model) --> Can Null Hypothesis be rejected for them? What is the role the of Pearson chi2-test value of 1.16 in this case with df 18? Or the log likelihood of -13669. ? All help would be very appreciated!

Generalized Linear Model Regression Results
Dep. Variable: y1,y2
No. Observations: 317
Df Residuals: 298
Model: GLM Model
Family: Binomial
Df Model: 18
Link Function: logit Scale: 1.0
Method: IRLS Log-Likelihood: -13669.
Deviance: 24554.
Pearson chi2: 1.16
No. Iterations: 7

var, coef, std err, z, P>|z|, [0.025, 0.975]

x1 0.0016, 1.91E-05, 83.485, 0.000, 0.002, 0.002

x2 3,07E-02, 3.16E-07, 96.862, 0.000, 3.00E-05, 3.13E-05

x3 -25.661, 0.008, -332.962, 0.000, -2.581, -2.551

x4 -20.821, 0.008, -251.433, 0.000, -2.098, -2.066

..

x22
 
Last edited: