Games (1973) criticized the Marascuilo and Levin (1970) "nested" or simple effects design as a poor substitute for the usual logic of a factorial analysis of variance. Marascuilo and Levin's use of "nesting" is completely idiosyncratic and contrary to well-established usage, thus confusing the student. Proper usage is illustrated. Statistical usage modifies models to fit the reality of the data, just as scientific models are modified to fit the facts. All present scientific models are wrong. We seek gradual improvements so our models come ever closer to the truth, but we recognize them as approximations with a certain margin of error. Similarly in data analysis we search for a simple model that describes the data within the limits of sampling error. Texts and articles reflecting this approach are cited. Marascuilo and Levin (1976) actually claim that Type IV errors are as important as Type I and Type II errors. They advocate extreme rigidity as a virtue disguising it under the name of "elegance". The flexibility desired in a good scientist is equally desirable in a good data analyst. (Author)