STATISTICAL MODELING OF AGE TRENDS IN MASTERS ATHLETES
Research on age trends in Masters Athletes always seemed special to me for several reasons: high quality of measurement on familiar and highly practiced activities; performance trends not confounded by effects of chronic incapacity or physical inactivity; a rare opportunity to study expertise at the highest level. As someone involved in such research since near its inception in the 1970s, I’m delighted with the opportunity to revisit old discoveries, trace their fate over the ensuing decades, and try to introduce new methodology and findings.
A brief reminiscence might be a good way to begin because the process of discovery has relevance to its outcomes. Three major changes from my earliest research until now are significant: these relate to accessibility of data, quality of record performances, and statistical methodology. First, records then were not easily accessible; one compilation I tracked down was available only on mimeographed sheets. International bodies such as World Masters Athletics (WMA) and the Fédération International de Natation (FINA) now put world records for track and field and swimming on the Internet for anyone to peruse. Second, the quality of performance by Masters Athletes is much higher now than it was then because of greater participation and more opportunities for competition. Third, for analysis then I used a HP55 calculator that was hightech for the era. It had four built-in curve-fitting models that seemed so sophisticated. Statistical procedures such as mixed linear analysis, essential for analysis of nested data, were not even a dream in some developer’s eye. Consequently, analysis now compared with then should be more sophisticated and based on more readily accessible data of higher quality.