Meeting of the Accreditation Committee

Jan. 30, 2017

Attending: Herman Marshall, Bob Ferrara, John Covert, Steve Carhart, Liz Jason, Brad Badgley, Kayla Lemay

0- Minutes review
	The previous minutes (Jan 9, 2017) were reviewed, corrected, and accepted.

1- Recent reviews
	All reports have been sent to the Dean except two.

2- Upcoming reviews
	Twelve groups will be reviewed this semester.  Nine have responded and all got their first choices of dates.  We will ensure that a member of the Accreditation committee will attend each day of reviews so that a proper overview may be given.  Herman will attend on Feb. 25, Bob on March 18, and Steve on April 29.  Only one team is needed on Feb. 25, and we have sufficient volunteers.  (Pam Gannon was later added to the team for the morning session, her first time on a review team.)

3- Annual report
	No progress.

4- BDF privacy and aggregation for AILG use
	Tom Yu, Herman, and John met on Jan. 23 to discuss this issue.  A method was worked out that should satisfy privacy issues as well as provide the desired information.
	The procedure starts with a request for BDF information.  A requesting group specifies the needed criteria for participation (by fraction of living groups in a category or by total number, etc.).  The group would generate fake BDFs with known data and an aggregation algorithm so that the results of aggregation of data will be known.  The Accreditation Coordinator (AC, John Covert) processes these BDFs to provide the aggregated information that the requesting group verifies.  Once validated, the AC then verifies that the participation criteria are met for a specific category of living group before operating the aggregation algorithm on the relevant BDFs.  The results are then sent to the requestors.
	Several issues with any aggregation scheme were discussed.  The most difficult to handle is the one of missing data as not all living groups will opt-in to release BDF data. It has to be clear that information cannot be reverse-engineered from the results as to the missing data for those groups which did not opt-in, defeating the purpose of aggregation.  For example, if 25 of 26 fraternities agree to provide data and it is known which fraternity is not included, then it may be possible (but perhaps very difficult) to determine the missing data from the excluded fraternity. Another complication arises from reviews that are staggered throughout any given 2-year cycle.  For some groups, there are 2 or more reviews in two years; rarely, a review is put off for one cycle and in a few cases, a group was suspended or otherwise not operational for a two year cycle.  These are known 'corner cases' but there could be others that the aggregation algorithm should clearly handle.
	At the time of the meeting, it was known that the most recent request would not clear the ILG criterion for data inclusion (100% opt-in) and it was not clear if the criterion for fraternities would be satisfied, so further discussion was postponed to a later meeting, planned for 1/30 but was cancelled.

5- RPI report on 'Statement of Relationship with Greeks'
	Bob circulated some documents about RPI's agreement with Greek groups.  These are rather formal, so they might not be appropriate for adaptation to MIT.  At RPI, these groups have somewhat lower visibility on campus and cooperation with the administration.

6- Phone numbers in BDF
	We talked briefly about whether to add a box or two for phone numbers for both undergraduate and alumni groups.  It was deemed important as a way for the city to contact a living group but perhaps not enough reason to add to the BDF.  The AILG's IT committee is interested in this topic.

7- Rubric status
	We decided not to scrap it, yet.  Some more outreach about its purpose may be needed.

8- FSILG office participation in reviews
	No action at this time.

The next meeting will be on Mar. 7 at 3PM in W59.