ON A GROWTH PREDICTION OF JAPANESE HEIGHT

When only one physical observation is available at the time point of prediction, we propose a trial approach of growth prediction of height and obtain an expedient confidence region of predicted individual growth. We compare the proposed approach with an empirical Bayesian approach for future growth prediction from the statistical viewpoints of confidence region and goodness of prediction. This proposed approach is better than an empirical Bayesian approach although its expedient prediction is simple.