If I'm understanding things correctly, many of the institutional demographic characteristics are based on most recent award-year FAFSA data. What is the source of demographic data for students within a given cohort? As I understand it, the earnings cohorts are constructed based on entry year, even though all included students may not receive aid (or submit a FAFSA form) in that entry year. Earnings data is disaggregated by family income. Is this income based on a student's first submitted FAFSA form, an average of all submitted FAFSA forms for each student in the cohort, or something else?
Please note that the Scorecard documentation will answer these types of questions (available at https://collegescorecard.ed.gov/data/documentation/). In short, FAFSA data is not used, per se, but NSLDS data are used therefore you are correct that the earnings and debt cohorts are limited to students receiving federal financial aid. Other demographics (such as number of degree-seeking students enrolled) come from IPEDS and represent the entire institution, not just federally-aided students. The aid by family income data come from IPEDS as submitted by the institutions, and represents all first-time full-time students (not just federally-aided students).