I do not understand why the values of certain variables changed from the previous version to the current version. For example, grad_debt_mdn for the first observation (unitid=100654) is 25106 in MERGED_2010-11 datafile in the current version (updated on Sept 13, 2016). However, grad_debt_mdn for the first observation (unitidi=100654) is 27070.5 in merged_2010_PP in the previous version (updated on March 2, 2016). Given that the variable grad_debt_mdn from Merged_2010-11 measures NSLDS AY2009-10, AY2010-11 pooled cohorts, why should the value for grad_debt_mdn change when the data was updated? I did the checking and it was not caused by the inflation adjustment. I noticed this change also happened to a lot of other variables such as FIRSTGEN_YR6_N.

I thought the only update was adding the most recent data file "MERGED_2014-15 datafile", but apparently it is not the case. So how was the data updated? Is there a document that can show me the things you did to update the data?

  • I really like this question (and have upvoted it) because it gives absolutely no clue as to what you're talking about, and addresses the forum as though it were responsible for making these changes. I had to google "grad_debt_mdn" to figure out what this was about. Tony, you may have better luck sending this as an email to those who actually do control and distribute this data.
    – user3856
    Oct 7, 2016 at 14:35

1 Answer 1


In the previous version of Scorecard, the median undergraduate loan debt was calculated for those who separated (i.e., either graduated or withdrew) in a given fiscal year , measured at the time of calculation (October 2014). These values were recalculated and replaced, by calculating median debt for those who separate in a given fiscal year, measured at the time of separation. See the change log https://collegescorecard.ed.gov/data/changelog/ for more details and explanations of other changes made to previously published data values (note updates included changes beyond adding the 14-15 merged file).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.