ABSTRACT

Abraham Maslow (1966), writing on the psychology of science, noted that, "If the only tool

you have is a hammer, you tend to treat everything as if it were a nail." The corollary of this

proposition is that sometimes we drive nails, and sometimes we smash windows. The

statistical tools that we use to analyze data each carry with them certain assumptions about the

characteristics of the data. We use these methods so often that we can easily forget these

assumptions. Frequently, we have concentrated so much on teaching how to use these

methods that we may not have made clear the assumptions on which these methods are based,

or we may not even know those assumptions. When the data satisfy these assumptions, then

the analysis that we perform can provide faithful models of the phenomena that we wish to

describe. Otherwise, the analysis may be worthless, or dangerous, and worse yet, we may not

even question our methods. Like a chess player, we must look not only at the piece our

opponent just moved, but also remain aware of the move's strategic implications. We can

easily forget that a method as simple, as commonplace, as seemingly innocuous, as taking the

914 LIEBOVITCH ET AL.