Corrections
On page 91 line 13 of the first full paragraph, the 2.02 t value should be -2.02. Later on this page, the p value for the last line of the paragraph is .004, not > .001.
On page 97, we say that within SAS one cannot force the error variance to equal zero. It is possible though to use the "PARMS" command to force the error variance to a very small value, say 0.000001, which can accomplish the same purpose. The syntax for this is contained below in the SAS syntax in Data and Files on this page.
On page 97, the two-intercept model with heterogeneous variances, the fixed effects include Culture, Contribution, and their interaction. A more reasonable model would include sex as a fixed variable. Thanks to Joseph Cappelleri for pointing out this other problems in this chapter.
