Abstract
The authors propose a new nonparametric diagnostic test for checking the constancy of the conditional variance function σ2(x) in the regression model Yi = m(xi) + σ(x i)εi, i = 1,..., m. Their test, which does not assume a known parametric form for the conditional mean function m(x), is inspired by a recent asymptotic theory in the analysis of variance when the number of factor levels is large. The authors demonstrate through simulations the good finite-sample properties of the test and illustrate its use in a study on the effect of drug utilization on health care costs.
Original language | English (US) |
---|---|
Pages (from-to) | 545-558 |
Number of pages | 14 |
Journal | Canadian Journal of Statistics |
Volume | 33 |
Issue number | 4 |
DOIs | |
State | Published - Dec 2005 |
Keywords
- Analysis of variance
- Constant variance
- Health care costs
- Heteroscedastic errors
- Nearest-neighbour windows
- Nonparametric regression
- Residual