Free Trial

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

Share this Page URL
Help

11.11 ANCOVA > 11.11.4 R Functions ancdes, ancovamp, ancmppb, and ancmg - Pg. 619

Chapter 11 More Regression Methods 619 11.11.4 R Functions ancdes, ancovamp, ancmppb, and ancmg The R function ancovamp(x1,y1,x2,y2,fr1=1,fr2=1,tr=0.2,alpha=0.05,pts=NA) compares two groups based on trimmed means and takes into account multiple covariates using the method outlined in Section 11.11.3. The arguments are the same as ancova in Section 11.11.2. By default, pts=NA meaning that the points among the covariates at which the groups will be compared are determined by the function; it finds a point among the x1 values that has the deepest halfspace depth, plus the points on the 0.5 depth contour, and the groups are compared at these points provided that the corresponding sample sizes are at least 10. Should one want to pool the data and then find the deepest point, plus the points on the 0.5 depth contour, this can be done as indicated by an Example later in this section. The function controls the familywise error rate by determining a critical value via the Studentized maximum modulus distribution (as described in Chapter 7). The function ancmppb(x1,y1,x2,y2,fr1=1,fr2=1,tr=0.2,alpha=0.05,pts=NA,est=tmean,