Assessing the model fit

Description

Checking the fit of dose-response model by means of formal significance tests or graphical procedures.

Usage

1
  modelFit(object, test = NULL, method = c("gof", "cum"))

Arguments

object

object of class 'drc'

test

character string defining the test method to apply

method

character string specifying the method to be used for assessing the model fit

Details

Currently two methods are available. For continuous data the clasical lack-of-fit test is applied (Bates and Watts, 1988). The test compares the dose-response model to a more general ANOVA model using an approximate F-test. For quantal data the crude goodness-of-fit test based on Pearson's statistic is used.

None of these tests are very powerful. A significant test result is more alarming than a non-significant one.

Value

An object of class 'anova' which will be displayed in much the same way as an ordinary ANOVA table.

Author(s)

Christian Ritz

References

Bates, D. M. and Watts, D. G. (1988) Nonlinear Regression Analysis and Its Applications, New York: Wiley \& Sons (pp. 103–104).

Examples

1
2
3
4
5
## Comparing the four-parameter log-logistic model 
##  to a one-way ANOVA model using an approximate F test
## in other words applying a lack-of-fit test
ryegrass.m1 <- drm(rootl ~ conc, data = ryegrass, fct = W1.4())
modelFit(ryegrass.m1)   

Want to suggest features or report bugs for rdrr.io? Use the GitHub issue tracker.