). Based on “a multi-stage area probability sample that gives equal chance of selection to all non-institutional dwelling units in the conterminous United States [and that] was clustered geographically at each stage and stratified with interlaced controls,” it is a specimen of what Charles Lamb called biblia abiblia—things that have the outward appearance of books but are not books, since they cannot be read. Methodologically, it employs something called the “multivariate analysis,” which is explained in Appendix E. Typographically, Appendix E looks like language, but it turns out to be strewn with booby traps, all doubtless well known in the trade, like “dummy variables,” “F ratios,” “regression coefficients,” “beta coefficients” (and “partial beta coefficients”), and two kinds of “standard deviations”—“of explanatory variable A” and “of the dependent variable.”
My experience with such works may be summarized as follows: (alpha) the coefficient of comprehensibility decreases in direct ratio to the increase in length, or the longer the incomprehensibler, a notion that is illustrated here by the fact that Dr. Kolko’s short work is more understandable than Dr. Morgan et al.’s long one; (beta) the standard deviation from truism is inversely related to the magnitude of the generalization, or the bigger the statement the more obvious. (Beta) is illustrated by the authors’ five general proposals for action (“Implications for Public Policy”). The second of these is: “Fuller employment and the elimination of discrimination based on prejudice would contribute greatly to the independence of non-white persons, women, teenagers, and some of the aged.” That is, if Negroes and the rest had jobs and were not discriminated against, they would be better off—a point that doesn’t need to be argued or, for that matter, stated. The authors have achieved such a mastery of truism that they sometimes achieve the same monumental effect even in non-magnitudinous statements, as: “Table 28-1 shows that the proportion of parents who indicated that their children will attend private colleges is approximately twice as large for those with incomes over $10,000 as for those with incomes under $3,000.” Could be.
This is an excerpt from a 1963 article, sadly things seem to have only got worse in the meantime.
Comments