An analysis of longitudinal data augmented by cross-sectional data : A Bayesian approach

概要

論文の詳細を見る
It is often observed that whenever a longitudinal study on the developmental psychology or the clinical psychology is planned, obtaining the large sample size is so costful and prohibited. If the sample size of the longitudinal data is samll, the estimation on several parameters should tend to be unstable since the error variances are large. On the other hand, a cross-sectional study with large sample size is easily obtainable, although such data do not provide the repeated measurements. Therefore, if both a longitudinal data and a cross-sectional data are obtained from the same population and if these data are combined all together, the precision and the practical efficiency should be more satisfactory for the statistical analysis. In this paper, a longitudinal data and a cross-sectional data are assumed to be pooled together and the effect of the addition of the cross-sectional data to the longitudinal data is studied.
日本行動計量学会の論文
2002-08-00