Test for Varying Dispersion Parameter in Normal Linear Models
Performs Rao's score test for varying dispersion parameter in weighted and unweighted normal linear models.
## S3 method for class 'lm' vdtest(model, varformula, verbose = TRUE, ...)
model |
an object of the class lm which is obtained from the fit of a weighted or unweighted normal linear model. |
varformula |
an (optional) |
verbose |
an (optional) logical switch indicating if should the report of results be printed. By default, |
... |
further arguments passed to or from other methods. |
If the object model corresponds to an unweighted normal linear model then this test assess the assumption of constant variance, which
coincides with the (non-studentized) Breusch-Pagan test against heteroskedasticity.
A list which includes the three main attributes of the test for varying dispersion parameter: statistic ("statistic"), degrees of freedom ("df") and p-value ("p.value").
Breusch T.S. and Pagan A.R. (1979) A simple test for heteroscedasticity and random coefficient variation. Econometrica 47, 1287–1294.
Cook R.D. and Weisberg S. (1983) Diagnostics for heteroscedasticity in regression. Biometrika 70, 1–10.
## Example 1 fit1 <- lm(mpg ~ log(hp) + log(wt), data=mtcars) vdtest(fit1) ## Example 2 fit2 <- lm(Species ~ Biomass + pH, data=richness) vdtest(fit2) fit2a <- lm(Species ~ Biomass + pH, data=richness, subset=-c(1,3,18,20)) vdtest(fit2a) ## Example 3 whiteside <- MASS::whiteside fit3 <- lm(Gas ~ Temp + Insul + Temp*Insul, data=whiteside) vdtest(fit3) fit3a <- lm(Gas ~ Temp + Insul + Temp*Insul, data=whiteside, subset=-c(8,9,36,46,55)) vdtest(fit3a)
Please choose more modern alternatives, such as Google Chrome or Mozilla Firefox.