22 – Statistical Analysis of Women's Sports
Senior Swim Competition Times
David Doane
Oakland University
Kevin Murphy
Oakland University
This paper utilizes 565 observations on 500-yard freestyle swim times in the biennial U.S. National Senior Games (ages 50 and up) for 2009, 2011, and 2013. Five topics are discussed: (1) What is the effect of age on swim times, and how should this be modeled? (2) Do men and women exhibit the same patterns of decline by age? (3) Do years differ, say, as a result of stiffening competition? (4) What is the pattern of split times in this event? (5) Do starting block reaction times vary by age or gender? Three regression models are tried for the age-time relationship: age category binaries, quadratic, and semi-log. Estimation is by OLS and quantile regression, both with and without bootstrap standard errors. Similar age-time patterns emerge, though quantile regression helps to mitigate high extremes. Women's average times are slower, and their rate of decline may be slightly steeper than men's. There is evidence that women's times improved in successive biennial competitions. Split times are faster at the start and end of the race. Reaction times do not differ by gender, and only slightly by age.