var.levels <- expand.grid(gen=c("male", "female"),
Info=c("Support","Oppose"),
Health= c("Support","Oppose"))
( AIDS3way <- data.frame(var.levels, count=c(76, 114, 6, 11,
160, 181, 25, 48)) )
## gen Info Health count
## 1 male Support Support 76
## 2 female Support Support 114
## 3 male Oppose Support 6
## 4 female Oppose Support 11
## 5 male Support Oppose 160
## 6 female Support Oppose 181
## 7 male Oppose Oppose 25
## 8 female Oppose Oppose 48
# log linear
summary(mod1 <- glm (count ~ gen + Info + Health + gen*Health
+ gen*Info + Info*Health,
data=AIDS3way, family=poisson) )
##
## Call:
## glm(formula = count ~ gen + Info + Health + gen * Health + gen *
## Info + Info * Health, family = poisson, data = AIDS3way)
##
## Deviance Residuals:
## 1 2 3 4 5 6 7
## -0.10362 0.08516 0.39073 -0.26626 0.07183 -0.06730 -0.17923
## 8
## 0.13173
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 4.3426 0.1120 38.763 < 2e-16 ***
## genfemale 0.3856 0.1434 2.689 0.00717 **
## InfoOppose -2.7147 0.3035 -8.945 < 2e-16 ***
## HealthOppose 0.7269 0.1353 5.374 7.68e-08 ***
## genfemale:HealthOppose -0.2516 0.1749 -1.438 0.15035
## genfemale:InfoOppose 0.4636 0.2406 1.927 0.05401 .
## InfoOppose:HealthOppose 0.8997 0.2852 3.155 0.00160 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for poisson family taken to be 1)
##
## Null deviance: 445.82335 on 7 degrees of freedom
## Residual deviance: 0.30072 on 1 degrees of freedom
## AIC: 59.683
##
## Number of Fisher Scoring iterations: 4