Load packages and data
library(effects) ## load the effects package
library(car) ## for Anova: type II tests
data(Cowles, package = "carData")
Main effects model
mod.cowles0 <- glm(volunteer ~ sex + neuroticism + extraversion,
data=Cowles, family=binomial)
summary(mod.cowles0)
##
## Call:
## glm(formula = volunteer ~ sex + neuroticism + extraversion, family = binomial,
## data = Cowles)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.398 -1.045 -0.908 1.260 1.685
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -1.11650 0.24906 -4.48 7.4e-06 ***
## sexmale -0.23516 0.11118 -2.12 0.034 *
## neuroticism 0.00636 0.01136 0.56 0.575
## extraversion 0.06633 0.01426 4.65 3.3e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 1933.5 on 1420 degrees of freedom
## Residual deviance: 1906.1 on 1417 degrees of freedom
## AIC: 1914
##
## Number of Fisher Scoring iterations: 4
Anova(mod.cowles0)
## Analysis of Deviance Table (Type II tests)
##
## Response: volunteer
## LR Chisq Df Pr(>Chisq)
## sex 4.49 1 0.034 *
## neuroticism 0.31 1 0.575
## extraversion 22.14 1 2.5e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Test all interactions
mod.cowles1 <- glm(volunteer ~ (sex + neuroticism + extraversion)^2,
data=Cowles, family=binomial)
summary(mod.cowles1)
##
## Call:
## glm(formula = volunteer ~ (sex + neuroticism + extraversion)^2,
## family = binomial, data = Cowles)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.498 -1.054 -0.896 1.260 1.967
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -2.49818 0.57555 -4.34 1.4e-05 ***
## sexmale -0.02198 0.49187 -0.04 0.9644
## neuroticism 0.11545 0.03937 2.93 0.0034 **
## extraversion 0.17650 0.04273 4.13 3.6e-05 ***
## sexmale:neuroticism -0.00368 0.02302 -0.16 0.8730
## sexmale:extraversion -0.01466 0.02938 -0.50 0.6178
## neuroticism:extraversion -0.00881 0.00299 -2.95 0.0032 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 1933.5 on 1420 degrees of freedom
## Residual deviance: 1897.2 on 1414 degrees of freedom
## AIC: 1911
##
## Number of Fisher Scoring iterations: 4
Anova(mod.cowles1)
## Analysis of Deviance Table (Type II tests)
##
## Response: volunteer
## LR Chisq Df Pr(>Chisq)
## sex 4.92 1 0.027 *
## neuroticism 0.31 1 0.575
## extraversion 22.10 1 2.6e-06 ***
## sex:neuroticism 0.03 1 0.873
## sex:extraversion 0.25 1 0.618
## neuroticism:extraversion 8.81 1 0.003 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
Add one interaction
mod.cowles <- glm(volunteer ~ sex + neuroticism * extraversion,
data=Cowles, family=binomial)
summary(mod.cowles)
##
## Call:
## glm(formula = volunteer ~ sex + neuroticism * extraversion, family = binomial,
## data = Cowles)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.475 -1.060 -0.893 1.261 1.998
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -2.35821 0.50132 -4.70 2.6e-06 ***
## sexmale -0.24715 0.11163 -2.21 0.0268 *
## neuroticism 0.11078 0.03765 2.94 0.0033 **
## extraversion 0.16682 0.03772 4.42 9.7e-06 ***
## neuroticism:extraversion -0.00855 0.00293 -2.92 0.0036 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 1933.5 on 1420 degrees of freedom
## Residual deviance: 1897.4 on 1416 degrees of freedom
## AIC: 1907
##
## Number of Fisher Scoring iterations: 4
anova(mod.cowles0, mod.cowles, mod.cowles1)
## Analysis of Deviance Table
##
## Model 1: volunteer ~ sex + neuroticism + extraversion
## Model 2: volunteer ~ sex + neuroticism * extraversion
## Model 3: volunteer ~ (sex + neuroticism + extraversion)^2
## Resid. Df Resid. Dev Df Deviance
## 1 1417 1906
## 2 1416 1897 1 8.62
## 3 1414 1897 2 0.26
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