1993 Volume 20 Issue 1 Pages 77-90
This paper discusses macro analysis of social change by means of cohort analysis. The material is based on over thirty-five years of experience with Chikio Hayashi, conducting statistical analyses of the Japanese National Character studies (Hayashi, 1987; Hayashi and Suzuki, 1984; Suzuki, 1970). In the mid 1980s, a new Bayesian approach to cohort analysis was introduced (Nakamura, 1982, 1986). The details of this method are discussed elsewhere (see Sasaki and Suzuki, 1987; Glenn, 1989; Sasaki and Suzuki, 1989). We have used this new methodology, which can automatically partition age effects, period effects, and cohort effects, and here we will present some of our results using Bayesian cohort analysis (Suzuki, 1986).