rwm1984: rwm1984
In COUNT: Functions, Data and Code for Count Data

Description Usage Format Details Source References Examples

German health registry for the year 1984.

1	data(rwm1984)

A data frame with 3,874 observations on the following 17 variables.

docvis: number of visits to doctor during year (0-121)
hospvis: number of days in hospital during year (0-51)

edlevel: educational level (categorical: 1-4)
age: age: 25-64
outwork: out of work=1; 0=working
female: female=1; 0=male
married: married=1; 0=not married
kids: have children=1; no children=0
hhninc: household yearly income in marks (in Marks)
educ: years of formal education (7-18)
self: self-employed=1; not self employed=0
edlevel1: (1/0) not high school graduate
edlevel2: (1/0) high school graduate
edlevel3: (1/0) university/college
edlevel4: (1/0) graduate school

rwm1984 is saved as a data frame. Count models typically use docvis as response variable. 0 counts are included

German Health Reform Registry, year=1984, in Hilbe and Greene (2007)

Hilbe, Joseph, M (2014), Modeling Count Data, Cambridge University Press Hilbe, Joseph M (2011), Negative Binomial Regression, Cambridge University Press Hilbe, J. and W. Greene (2008). Count Response Regression Models, in ed. C.R. Rao, J.P Miller, and D.C. Rao, Epidemiology and Medical Statistics, Elsevier Handbook of Statistics Series. London, UK: Elsevier.

library(MASS)
library(msme)
data(rwm1984)

glmrp <- glm(docvis ~ outwork + female + age + factor(edlevel), family=poisson, data=rwm1984)
summary(glmrp)
exp(coef(glmrp))

summary(nb2 <- nbinomial(docvis ~ outwork + female + age + factor(edlevel), data=rwm1984))
exp(coef(nb2))

summary(glmrnb <- glm.nb(docvis ~ outwork + female + age + factor(edlevel), data=rwm1984))
exp(coef(glmrnb))

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

Call:
glm(formula = docvis ~ outwork + female + age + factor(edlevel), 
    family = poisson, data = rwm1984)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-3.5746  -2.1973  -1.2704   0.3265  26.0734  

Coefficients:
                  Estimate Std. Error z value Pr(>|z|)    
(Intercept)      -0.085762   0.041688  -2.057   0.0397 *  
outwork           0.258678   0.021532  12.013  < 2e-16 ***
female            0.273051   0.021185  12.889  < 2e-16 ***
age               0.022010   0.000851  25.863  < 2e-16 ***
factor(edlevel)2 -0.068450   0.042271  -1.619   0.1054    
factor(edlevel)3 -0.172508   0.039740  -4.341 1.42e-05 ***
factor(edlevel)4 -0.251967   0.048019  -5.247 1.54e-07 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

(Dispersion parameter for poisson family taken to be 1)

    Null deviance: 25791  on 3873  degrees of freedom
Residual deviance: 23957  on 3867  degrees of freedom
AIC: 31053

Number of Fisher Scoring iterations: 6

     (Intercept)          outwork           female              age 
       0.9178130        1.2952162        1.3139678        1.0222535 
factor(edlevel)2 factor(edlevel)3 factor(edlevel)4 
       0.9338402        0.8415515        0.7772701 

Call:
nbinomial(formula1 = docvis ~ outwork + female + age + factor(edlevel), 
    data = rwm1984)

Deviance Residuals:
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
-1.5676 -1.2638 -0.4750 -0.4535  0.1200  5.7030 

Pearson Residuals:
   Min. 1st Qu.  Median    Mean 3rd Qu.    Max. 
-0.6444 -0.6082 -0.3688 -0.0002  0.1273 26.9409 

Coefficients (all in linear predictor):
                 Estimate      SE      Z         p     LCL     UCL
(Intercept)       -0.1184 0.11089 -1.067     0.286 -0.3357  0.0990
outwork            0.2759 0.05949  4.639  3.51e-06  0.1593  0.3925
female             0.3252 0.05631  5.775   7.7e-09  0.2148  0.4356
age                0.0221 0.00237  9.322  1.14e-20  0.0174  0.0267
factor(edlevel)2  -0.1136 0.11822 -0.961     0.337 -0.3453  0.1181
factor(edlevel)3  -0.1758 0.10204 -1.723     0.085 -0.3758  0.0242
factor(edlevel)4  -0.3253 0.11491 -2.831   0.00465 -0.5505 -0.1000
(Intercept)_s      2.2582 0.07007 32.228 7.06e-228  2.1208  2.3955

Null deviance: 4148.15  on  3872 d.f.
Residual deviance: 3909.534  on  3866 d.f.
Null Pearson: 5923.495  on  3872 d.f.
Residual Pearson: 5703.044  on  3866 d.f.
Dispersion: 1.475179
AIC:  16637.37

Number of optimizer iterations:  84 

Warning messages:
1: In dnbinom(y, size = scale, mu = y.hat, log = TRUE) : NaNs produced
2: In dnbinom(y, size = scale, mu = y.hat, log = TRUE) : NaNs produced
3: In dnbinom(y, size = scale, mu = y.hat, log = TRUE) : NaNs produced
4: In dnbinom(y, size = scale, mu = y.hat, log = TRUE) : NaNs produced
5: In dnbinom(y, size = scale, mu = y.hat, log = TRUE) : NaNs produced
6: In dnbinom(y, size = scale, mu = y.hat, log = TRUE) : NaNs produced
7: In dnbinom(y, size = scale, mu = y.hat, log = TRUE) : NaNs produced
     (Intercept)          outwork           female              age 
       0.8883717        1.3177478        1.3843133        1.0223234 
factor(edlevel)2 factor(edlevel)3 factor(edlevel)4    (Intercept)_s 
       0.8926237        0.8388077        0.7223425        9.5654586 

Call:
glm.nb(formula = docvis ~ outwork + female + age + factor(edlevel), 
    data = rwm1984, init.theta = 0.442880598, link = log)

Deviance Residuals: 
    Min       1Q   Median       3Q      Max  
-1.5672  -1.2639  -0.4748   0.1198   5.6998  

Coefficients:
                  Estimate Std. Error z value Pr(>|z|)    
(Intercept)      -0.113599   0.111888  -1.015  0.30997    
outwork           0.276468   0.062895   4.396 1.10e-05 ***
female            0.325059   0.059812   5.435 5.49e-08 ***
age               0.021969   0.002404   9.138  < 2e-16 ***
factor(edlevel)2 -0.114024   0.118122  -0.965  0.33439    
factor(edlevel)3 -0.176290   0.102102  -1.727  0.08424 .  
factor(edlevel)4 -0.325601   0.115847  -2.811  0.00494 ** 
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

(Dispersion parameter for Negative Binomial(0.4429) family taken to be 1)

    Null deviance: 4148.4  on 3873  degrees of freedom
Residual deviance: 3909.8  on 3867  degrees of freedom
AIC: 16637

Number of Fisher Scoring iterations: 1


              Theta:  0.4429 
          Std. Err.:  0.0137 

 2 x log-likelihood:  -16621.3670 
     (Intercept)          outwork           female              age 
       0.8926162        1.3184649        1.3841121        1.0222123 
factor(edlevel)2 factor(edlevel)3 factor(edlevel)4 
       0.8922363        0.8383752        0.7220934