The GSS-NDI is a very complex dataset to analyze. It has data from 1978-2010 linked to mortality follow up through 2014. If you just run simple survival analyses, people surveyed in 1978 will have much greater follow up than those surveyed in 2010. As a result, you need to be very careful how you code for survival time. You also need to be very careful about cohort effects since 50-year-olds in 1978 are 3 generations removed from 50-year-olds in 2010. Research that has already been peer-reviewed and finished can be posted here to help those who might need help with coding. Annotated work is greatly appreciated.