Bootstrap and Asymptotic Prediction Criterion Estimates for Binomial Proportions in Insemination Data