Hi Statalist,
I am attempting to run a model on stata, specifically, a Variance function regression (VFR) embedded in a hierarchical Age-period-cohort (cross-classified multilevel model). I am looking at both between and within-group differences in American's mental health over various periods and cohorts. I am using a cross-classified multilevel model with the assumption that individuals are simultaneously nested in and cross-classified by two higher-level social contexts defined by birth cohort membership and time period. My data is cross-sectional and I have three different time periods and seven different cohorts.
This is the code I am using for my cross-classified model:
Code:
mixed Mental Age Nchildren i.financial i.Sex i.chronic i.Ses || _all: R.Period || Cohort:, mle variance
where Mental is a continuous outcome variable (a factor score)
Financial is a dummy for whether individual is experiencing financial stress
Chronic is a dummy for whether individual has chronic illness
The model runs fine and all results for all co-variates make sense.
However, I am trying to follow this paper:
Zheng, H., Yang, Y. and Land, K.C., 2011. Variance function regression in hierarchical age-period-cohort models: Applications to the study of self-reported health. American sociological review, 76(6), pp.955-983.
The two steps I would like to do are:
1. Estimate the B regression coefficient vectors for between group mental health across age, period and cohort
2. Estimate the λ regression coefficient vectors for within group mental health across age, period and cohort
(As the authors do on page 967)
I have checked the stata manual and various tutorials but I am unable to figure out the code to use the HAPC model to estimate the variance function regression model and thus separate my B coefficients and my λ coefficients. The intuition behind this is that I would like to analyze the sources of variations in American's mental health and the differential contributions of age, period and cohort to variations in mental health.
Any help would be greatly appreciated as I've hit a wall and I am a beginner at Stata.
Thank you,
Hale Isaac
Western, B. and Bloome, D., 2009. 9. Variance Function Regressions for Studying Inequality. Sociological Methodology, 39(1), pp.293-326.
