Here are the mistakes that have been caught so far.

1.  In Chapter 4, there is no blank line between the abstract and the chapter text.  The abstract should be just the very first paragraph, then the chapter begins with the next paragraph, which begins: "Systems and scale are the yin and yang of your study preliminaries." 

2.  In Chapter 5, there is a mistake in my discussion of the consequences of a variable being measured with error.  In the example at the bottom of p. 62, a binary variable, B, is mis-measured 2% of the time.  The text argues that this error is relatively inconsequential if this binary variable is the dependent variable, and that the only effect is a slight inflation of the standard errors.  This would be true if B were a continuous independent variable that was measured with random error, but it is not true when B is a binary variable, because any error in measurement is related to the true value of B.  Thus, there will also some bias (though the extent of bias is unclear).  Caught by Dan Millimet of Southern Methodist University.