Estimation of monotone function for data on records

In several regression problems monotonicity is a key feature of the underlying regression function. Although in some cases the observations are not strictly monotonic due to random error, in some other cases the observations may be monotonic by nature. In all such cases the fitted curve should catch this monotonicity in order to explain the dataset. In the present paper, the dataset on the development of the world record on men’s 100 m sports is considered for analysis. Using Bayesian methodology the fittings of the data is described by two methods, namely using monotone spline and the local regression technique of O’Hagan (1978). A Bayesian prediction for the future world record is also considered