affairs: affairs

In COUNT: Functions, Data and Code for Count Data

Description

Data from Fair (1978). Although Fair used a tobit model with the data, the outcome measure can be modeled as a count. In fact, Greene (2003) modeled it as Poisson, but given the amount of overdispersion in the data, employing a negative binomial model is an appropriate strategy. The data is stored in the affairs data set. Naffairs is the response variable, indicating the number of affairs reported by the participant in the past year.

Usage

data(affairs)

Format

A data frame with 601 observations on the following 18 variables.

naffairs

number of affairs within last year

kids

1=have children;0= no children

vryunhap

(1/0) very unhappily married

unhap

(1/0) unhappily married

avgmarr

(1/0) average married

hapavg

(1/0) happily married

vryhap

(1/0) very happily married

antirel

(1/0) anti religious

notrel

(1/0) not religious

slghtrel

(1/0) slightly religious

smerel

(1/0) somewhat religious

vryrel

(1/0) very religious

yrsmarr1

(1/0) >0.75 yrs

yrsmarr2

(1/0) >1.5 yrs

yrsmarr3

(1/0) >4.0 yrs

yrsmarr4

(1/0) >7.0 yrs

yrsmarr5

(1/0) >10.0 yrs

yrsmarr6

(1/0) >15.0 yrs

Details

rwm5yr is saved as a data frame. Count models use naffairs as response variable. 0 counts are included.

Source

Fair, R. (1978). A Theory of Extramarital Affairs, Journal of Political Economy, 86: 45-61. Greene, W.H. (2003). Econometric Analysis, Fifth Edition, New York: Macmillan.

References

Hilbe, Joseph M (2011), Negative Binomial Regression, Cambridge University Press Hilbe, Joseph M (2009), Logistic regression Models, Chapman & Hall/CRC

Examples

data(affairs)
glmaffp <- glm(naffairs ~ kids + yrsmarr2 + yrsmarr3 + yrsmarr4 + yrsmarr5,
               family = poisson, data = affairs)
summary(glmaffp)
exp(coef(glmaffp))

require(MASS)
glmaffnb <- glm.nb(naffairs ~ kids + yrsmarr2 + yrsmarr3 + yrsmarr4 + yrsmarr5,
                   data=affairs)
summary(glmaffnb)
exp(coef(glmaffnb))

Example output

Loading required package: msme
Loading required package: MASS
Loading required package: lattice
Loading required package: sandwich

Call:
glm(formula = naffairs ~ kids + yrsmarr2 + yrsmarr3 + yrsmarr4 + 
    yrsmarr5, family = poisson, data = affairs)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-1.9668  -1.9364  -1.5412  -0.9274   7.0799  

Coefficients:
            Estimate Std. Error z value Pr(>|z|)    
(Intercept)  0.34038    0.09182   3.707  0.00021 ***
kids         0.28809    0.09371   3.074  0.00211 ** 
yrsmarr2    -1.18431    0.17058  -6.943 3.84e-12 ***
yrsmarr3    -0.45650    0.10536  -4.333 1.47e-05 ***
yrsmarr4    -0.11823    0.09896  -1.195  0.23220    
yrsmarr5     0.03119    0.09912   0.315  0.75303    
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

(Dispersion parameter for poisson family taken to be 1)

    Null deviance: 2925.5  on 600  degrees of freedom
Residual deviance: 2797.0  on 595  degrees of freedom
AIC: 3303

Number of Fisher Scoring iterations: 7

(Intercept)        kids    yrsmarr2    yrsmarr3    yrsmarr4    yrsmarr5 
  1.4054793   1.3338755   0.3059569   0.6334955   0.8884915   1.0316802 

Call:
glm.nb(formula = naffairs ~ kids + yrsmarr2 + yrsmarr3 + yrsmarr4 + 
    yrsmarr5, data = affairs, init.theta = 0.1188516427, link = log)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-0.8253  -0.8155  -0.7611  -0.6002   1.9331  

Coefficients:
            Estimate Std. Error z value Pr(>|z|)   
(Intercept)  0.33180    0.31247   1.062  0.28830   
kids         0.27309    0.31104   0.878  0.37995   
yrsmarr2    -1.19445    0.42962  -2.780  0.00543 **
yrsmarr3    -0.38936    0.35449  -1.098  0.27205   
yrsmarr4    -0.08137    0.38166  -0.213  0.83116   
yrsmarr5     0.07220    0.40464   0.178  0.85838   
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

(Dispersion parameter for Negative Binomial(0.1189) family taken to be 1)

    Null deviance: 344.44  on 600  degrees of freedom
Residual deviance: 332.40  on 595  degrees of freedom
AIC: 1504.6

Number of Fisher Scoring iterations: 1


              Theta:  0.1189 
          Std. Err.:  0.0127 

 2 x log-likelihood:  -1490.6260 
(Intercept)        kids    yrsmarr2    yrsmarr3    yrsmarr4    yrsmarr5 
  1.3934701   1.3140193   0.3028717   0.6774934   0.9218498   1.0748732 

COUNT documentation built on May 2, 2019, 2:37 a.m.