[6413] in s-news-athena
Poisson Regression
daemon@ATHENA.MIT.EDU (YONG CHANG )
Wed Feb 8 12:18:23 1995
From: ychang@welchlink.welch.jhu.edu (YONG CHANG )
To: s-news@utstat.toronto.edu (S Newsgroup)
Date: Wed, 8 Feb 1995 11:30:36 -0500 (EST)
Is there anyone out there who has encountered an increase of residual
deviance over null deviance in poisson regression analysis?
The following is the script that I used to generate a simple poission
run. The second model with "yearrisk" variable has an increase of
residual deviance over null deviance, which is what really puzzles
me.
Thanks in advance!!!
-----------------------------------------------------------------
Call: glm(formula = death ~ agerisk + offset(log(pyrs)), family = poisson, data=
all.process[all.process$pyrs != 0, ])
Deviance Residuals:
Min 1Q Median 3Q Max
-3.822337 -1.302971 -0.4514509 0.9317032 5.001732
Coefficients:
Value Std. Error t value
(Intercept) -5.905073 0.02938412 -200.96141
ageriskz55-64 1.484744 0.03734658 39.75584
ageriskz65+ 2.583656 0.03449133 74.90740
Null Deviance: 5990.903 on 47 degrees of freedom
Residual Deviance: 157.0649 on 45 degrees of freedom
Number of Fisher Scoring Iterations: 3
-----------------------------------------------------------------
Call: glm(formula = death ~ yearrisk + offset(log(pyrs)), family = poisson,data=
all.process[all.process$pyrs != 0, ])
Deviance Residuals:
Min 1Q Median 3Q Max
-23.64872 -5.894786 1.507092 8.711542 29.31456
Coefficients:
Value Std. Error t value RR Lower Upper
(Intercept) -4.8680463 0.01851132 -262.97670 0.01 0.01 0.01
yearrisk 0.3676398 0.02552517 14.40303 1.44 1.37 1.52
(Dispersion Parameter for Poisson family taken to be 1 )
Null Deviance: 5990.903 on 47 degrees of freedom
Residual Deviance: 6571.964 on 46 degrees of freedom
Number of Fisher Scoring Iterations: 4
