aods3 (version 0.4-1.2)

gof: Test of Goodness-of-Fit of Models for Count data

Description

The function returns a chi-squared test of goodness of fit for models of class glm, aodml or aodql.

Usage

gof(object)
  gof.default(object)
  
  # S3 method for gof
print(x, ..., digits =  max(3, getOption("digits") - 3))

Value

An object of class gof, printed with print.gof.

Arguments

object

An object of class glm, aodml or aodquasi.

x

An object of class gof.

digits

A numerical scalar indicating the number of digits to be printed after the decimal place.

...

Further arguments passed to print.

Details

Function gof calculates the deviance $D$ and the Pearson chi-squared $X^2$ statistics for the model under consideration. Let $y$ be the observed response, and $E[y]=\mu$ and $Var[y]$ its mean and variance estimated from the model, statistic $X^2$ is calculated by:

$$X^2=\sum _i((y_i-\mu {)}^2/Var[y_i])$$

Assuming that the data length is $N$ and the number of the parameters in the model is $p$, eqnD and eqnX^2 are compared to a chi-squared distribution with $N-p$ degrees of freedom.

References

Agresti, A. Categorical data analysis. Wiley, 1990.

See Also

residuals, chisq.test

Examples

data(orob2)
fm1 <- glm(cbind(m, n - m) ~ seed, data = orob2, family = binomial)
fm2 <- aodml(cbind(m, n - m) ~ seed, data = orob2, family = "bb")
gof(fm1)
gof(fm2)