EMAIL GROVER |

I have often feared that, by adopting and ubiquitously using the term "standard deviation," statisticians are subtly promulgating the concept that there is a certain level of aberration or perversity that is somehow legitimized even inside a well-ordered universe like mathematics. "He was caught doing What else can be expected from statistics, a field that is quite darkly values-prescriptive, almost in a Faustian sense? In order to be "normal," you must lie within the confines of a very rigid mathematical formula, or else you're branded as somehow "skewed." Even if you do manage to shoehorn yourself into that narrow definition of normalcy, where can you go to be acceptable? If you are on the edges, you are an "outlier," suggesting fringe prevarication. And no matter where you try to fit in, your place in life is measured, not by how good or wholesome you are, but by your "variance"--which is to say, how many multiples of some measure of standard (acceptable) deviations (perversions) you are from the aptly-named "mean" center of your universe. Once you have found your place, don't think it gets any less diabolical. What else can be said of a system where inference, standard error, and bias are commonplace; where the Ten Commandments have been reduced merely to two types of error; and where power is never ascribed to the population, but rather to the rigorous tests to which they are subjected? Absolutes (all things resembling constants) are virtully ignored. Relativism is worshipped to the extent that only variables are observed, measured, and quantified; have their variances analyzed; and are given freedom (the degrees of which are carefully controlled). Only if your perspective comes from a critical region of your existence may you reject the prevailing nihilistic gestalt--and only by attempting to reject that null hypothesis do you discover that, as your confidence level goes up, the likelihood of your succeeding diminishes. Finally, your quest to satisfy the age-old human need to predict future outcomes--far from leading you towards progress--turns out only to be an exercise in regression. And I'm 95% confident that I'm right about all that. |