Abstract
In the paper we derive a test of fit of a linear model based on a nonparametic kernel estimator of a regression function. The test is especially useful when no replication is available.

This publication has 6 references indexed in Scilit: