Statistical tools and status symbols.

Abstract

This paper sets the stage for a series of edited papers which follow. It is argued that obvious statistical blunders (mortal sins) have not disappeared entirely from the aging literature, but that the most frequent error now is that of choosing a less powerful analytic solution when a more powerful (and equally applicable) one is at hand (a venial sin). Clearly, it is argued, power must be considered in relationship to the theoretical model, the condition of the data set, and the research question. Editors and reviewers, it is argued, must cut through the "glitter" of new fads and the security of tradition in order to objectively consider the merits of new statistical applications. More importantly, without knowledge of the issues, limitations, and advantages of contemporary and emerging analytic techniques, authors cannot make good analytic choices and defend them effectively. The time between presentation of innovative approaches in the quantitative journals and application in the aging literature contributes to traditionalism in data analysis. Overzealousness by advocates of new approaches often contributes to premature embracing of them. Thus the papers that follow were designed to bring new methods to the "working-researcher" with a frank appraisal of assumptions, proper and improper applications, and limitations. Authors of these papers were challenged to communicate in plain and understandable terms and to provide practical examples.