Another compelling reason for releasing the actual DMDC SAGR reports

We have repeatedly made the point (e.g., here and here) that the administration should release the actual Sexual Assault and Gender Relations (“SAGR”) survey reports from the Defense Manpower Data Center (‘DMDC”) instead of deigning to provide us only with its “spin” of those reports. The DMDC SAGR report for the other federal academies is public and there is no reason that the Academy’s report should not also be public. It is evident that there is something damning in those reports that the administration does not want us (or Congress) to see.

We previously pointed out that  the administration’s spin of the 2013-14 DMDC SAGR report primarily reports data in percentages, whereas the administration’s spin of the 2011-12 DMDC SAGR report set forth the data in numbers. The use of percentages makes it hard to understand the severity of the issues and the change in methodology adopted by the administration spinners makes it difficult to compare the Academy’s performance in addressing SA/SH.

We also have observed that the spun version of the 2013-14 of the SAGR report uses internally inconsistent definitions of “sexual harassment.” This error in interpretation is yet another reason that the underlying data should be released so that everyone can see what the experts at DMDC used as the definition rather than what the administration used when spinning the data for its own purposes.

Now we’ve identified yet another reason: The administration’s math doesn’t add up. In Table 1 of the spun version of the 2013-14 SAGR report (at page 13 of this pdf), the administration reports that 761 midshipmen responded to the survey and 537 completed it. On page 9 of that same pdf, the administration’s mathematically-challenged spinners stated that 663 students took the survey in April 2014 and 125 took it in August 2014. The spinners did not report the number of students who took the survey when the DMDC team returned in November 2014, but the April and August numbers alone sum up to 788 students – 27 more than the total number of students who took the survey.

Elsewhere in the report (page 12 of the pdf), the spinners state that “34 percent of  Midshipmen completed the survey in November” [2014]. (Emphasis added.) But, 34% of the 537 students who completed the survey equals 183; so when we add up the April, August and November numbers, we now have the administration spinners indicating that a minimum*** of 971 students took the survey when Table 1 shows only 761 midshipmen took the survey (and the total student population was 936 per the same table).

We cannot trust any of the information in the administration’s spin of the DMDC SAGR reports. The administration’s spin – which is what was given to Congress – is flawed and unreliable. How can anyone have confidence in any actions taken by this administration when it cannot (or chooses not to) summarize accurately a report prepared by the recognized experts in surveying SAGR at the federal academies?  How can Congress intelligently address SA/SH at the Academy when the document it must rely upon (and the administration) is so badly discredited? It is long past the time when the DMDC SAGR reports should have been released to the public.

___________

*** It’s a “minimum” number because it is based upon midshipmen who “completed” the survey.  Unless every single midshipman who took the survey in November 2014 also completed it, the number of midshipmen who took the survey in November 2014 would be greater than 183