Table of Contents#### Download Safari Books Online apps: Apple iOS | Android | BlackBerry

Entire Site

Free Trial

Safari Books Online is a digital library providing on-demand subscription access to thousands of learning resources.

584 Introduction to Robust Estimation and Hypothesis Testing Section 11.5.12 in conjunction with runm3d. (When y is binary, the function logadr fits a generalized additive model in conjunction with Copas's method previously described.) The R function gamplotINT(x, y, pyhat = F, plotit = T, theta = 50, phi = 25, expand = 0.5, scale = F, zscale = T, eout = F, outfun = out, ticktype = "simple",) is like gamplot, only it is limited to p = 2 predictors and is based on the model y = g 1 (x 1 ) + g 2 (x 2 ) + g 3 (x 1 , x 2 ) + rather than y = g 1 (x 1 ) + g 2 (x 2 ) + . This is useful when checking for interactions as described in Section 11.7. 11.6 Checking the Speci cation of a Regression Model Typically, when testing hypotheses, a particular parametric form for a regression model is specified and inferences are made about the parameters assuming that the model is correct. A practical concern is that the assumed parametric form might be wrong, which in turn can lead to erroneous conclusions. As a simple example, values for x were generated by the