Attendees:

Regrets:

It should be possible to construct an entitlement to enforce a policy if the convention is supported on both sides. Is this convention something that should be standardized? Depends on the scale you want to work, and it might not be feasible to generalize on.

Perhaps something like a RAiD to uniquely identify the project(s) someone is a PI for? https://www.raid.org.au/

A concern that the descriptions are too specific to the NIH. Need to reach out to other research grant authorities to see how the definitions are similar. Alternatively, if the naming convention is granting authority+value (e.g., NIH.gov_$value) then that scopes the value to the NIH’s definition. If you go with the eduPerson entitlement naming scheme, this idea is built in to the URN or even the URI.

The problem with entitlement is that it hasn’t scaled. While there is value when you can get it to work to scope things down as specific as needed, the real value of the entitlement-based approach is when you have enough generalities that you’ll get wider support for the value.
Go with the entitlement path to get some research community more heavily engaged, and build adoption off that.

Pre-registrant - will likely have a very local definition to the point of not being usable
Researcher - if you use IPEDs (in the US) then this might be something more usable
An affiliation is just the relationship. Some of the relationships we’re looking at depend on the regional context (e.g., pre-higher ed doesn’t make sense outside of a US context). Affiliation is not part of the identity lifecycle process.

Some affiliations should not be released during a fed authN workflow. E.g., former-student introduces FERPA concerns in the US (though they may still have access to online resources as a result of being a former-student)

Filtered attribute release is critical to all of this.
It does not help to clarify to people what’s wrong; you have to tell them what to do instead. “Here’s the entitlement use case you should use instead of using affiliation."
Consultant - perhaps Contractor would be better? Most places have differentiation between employees and contractors. This is almost a vendor, except there may be access required to some systems. Another option is to use the word “Temporary”. These may also be considered Sponsored accounts, which is a general enough term that it may apply across the board.

Sponsored: An identity that is being managed because a perceived business need for someone not otherwise affiliated.

Regarding the various student-related value (post-graduate vs graduate vs under-graduate) is splitting the students into various sub-categories a useful exercise? What problem are we trying to solve.
Maybe these local needs should be handled via groups instead of a schema?
Student-employee = this varies between institutions as well. In some definitions, this is a financial-aid defined role. In other organizations, this isn’t actually tied to financial aid. In The Ohio State example, employee trumps student-ness.

The two affiliations that have some measure of consensus are:

Group will consider whether there’s a better word than “sponsored” and discuss on the next call.

Next call: 14 July 2020