I noticed that for many CSD observations at chain schools, such as University of Phoenix, the earnings and repayment variables in levels (#repayers, #working 6 years after, etc.) seem disproportionately high, even after accounting for the fact that they are pooled cohorts, and I am assuming this is because this data is reported at the OPEID level. What OPEID or alternative variable is used to identify/merge in this information to the CSD? (i.e. 6-digit OPEID, 8-digit OPEID).
For the data elements available only for Title IV recipients (earnings, repayment rate, etc.), data produced for Scorecard are rolled up to the 6-digit OPEID level. Please note that the unit of analysis for Scorecard is the IPEDS UNITID. In cases where IPEDS UNITIDs share a common 6-digit OPEID, all observations are assigned the same (student-weighted) average outcome or median outcome for students across all UNITIDs with a common 6-digit OPEID.