BACKGROUND:Variation in the relationship between education and health has been studied intensely over the past few decades. Although there is research on gender disparity and cohort variations in educational effect on health using samples from the U.S. and Europe, research about China's is limited. Given the specific social changes in China, our study is designed to analyze the gender and cohort patterns in the education-health gradient. METHOD:The latent growth-curve modeling was used to analyze the gender and cohort variations in the education gradient in self-rated health among Chinese respondents. The study employed longitudinal and nationally representative data from the Chinese Family Panel Studies from the years 2010 to 2016. Each cohort is specified according to their distinct periods of social change in China. Following the analysis, we used latent growth-curve model to illustrate gender and cohort differences in the age-graded education and health trajectories. RESULTS:Although Chinese men have reported to have better health than women in general, women reported 1.6 percentage points higher in self-reported health for each additional year of schooling compared to that of men (P < 0.001). The latent growth curve model showed women's extra education benefits were persistent overtime. Compared to the people born during the "Old China" (1908-1938), the education gradient in self-rated health did not change for cohorts born before 1955 and after 1977, but the education-health gap changed significantly in the 1956-1960 (O.R. = 1.038, P < 0.05), 1967-1976 (O.R. = 1.058, P < 0.001), and 1977-1983 (O.R. = 1.063, P < 0.001) cohorts. There was a gender difference for the cohort variations in the education-health gradient. For women, the education effect in the 1956-1960 (O.R. = 1.063, P < 0.05), 1967-1976 (O.R. = 1.088, P < 0.001) and 1977-1983 (O.R. = 1.102, P < 0.001) cohorts was significantly higher than that of the 1908-1938 cohort. On the contrary, the education-health gradient remained the same across all cohorts for men. CONCLUSION:Our study suggests that the education-health gradient varies across cohorts for women, but the size of education effect remains consistent for men across cohorts. The findings support the resource-substitution hypothesis and not the rising-importance hypothesis in China. We discussed the potential influences of the unique, social transformation and educational expansion in China.

Studies have shown that happiness level varies significantly across birth cohorts and baby boomers are the unhappiest of all birth cohorts. Yet, we don’t know if this is due to their large cohort size negatively affecting happiness. We question whether people born in high fertility times are unhappy because they suffer more from economic setbacks and/or social strains. Using 9 waves of data from the European Social Survey United Kingdom Subset 2002–2018 (N = 19,364) and hierarchical age-period-cohort cross-classified models, we analyze the effects of cohort size, socioeconomic status, marital status, and sociality on happiness. Cohort size, marital status, and sociality are the top three factors of cohort difference in happiness, but socioeconomic status is not. Cohort size is negatively associated with happiness. Income, education, or employment are not the source of unhappiness among the Boomers. Besides being members of a large cohort, the Boomers have two known factors against their odds: they are the most likely to separate and divorce and the least likely to socialize with friends despite having a large number of peers. Social disintegration and deprivation, not economic impoverishment, appears to be the culprit of unhappiness of the UK Baby Boomers.

The interdisciplinary field of knowledge discovery and data mining emerged from a necessity of big data requiring new analytical methods beyond the traditional statistical approaches to discover new knowledge from the data mine. This emergent approach is a dialectic research process that is both deductive and inductive. The data mining approach automatically or semi-automatically considers a larger number of joint, interactive, and independent predictors to address causal heterogeneity and improve prediction. Instead of challenging the conventional model-building approach, it plays an important complementary role in improving model goodness of fit, revealing valid and significant hidden patterns in data, identifying nonlinear and non-additive effects, providing insights into data developments, methods, and theory, and enriching scientific discovery. Machine learning builds models and algorithms by learning and improving from data when the explicit model structure is unclear and algorithms with good performance are difficult to attain. The most recent development is to incorporate this new paradigm of predictive modeling with the classical approach of parameter estimation regressions to produce improved models that combine explanation and prediction.