Cohort, age, and period effects in the analysis of U.S. suicide patterns: 1933-1978

This study employed multiple classification analysis to examine the impact of age, period, and cohort effects on U.S. white male suicide rates between 1933 and 1978. Previous studies of cohort effects with regard to suicide behavior have been descriptive in nature, and have failed to consider the interaction of the three effects. Drawing on a study of pulmonary tuberculosis mortality by W.M. Mason and Smith (1985), the work specified seven models. The Shazam ordinary-least-squares computer package was employed to estimate the logit coefficients for the seven models. Period effects were found to be weaker than age and cohort effects for explaining shifts in white male suicide patterns. In this study all of the three effects were measured indirectly, and this fact limits the validity of the findings. Causal analysis, which directly measures at least one of the effects, would have improved the robustness of the findings.