Going forward, we will post new minutes and other TAG announcements here.
Identification Code Lookup
Situation:
A student can be identified by one of several different codes.
Submission of Ed-Fi data requires knowing the StudentUniqueId as stored for the given Ed-Fi API
A local installation might be using a different StudentUniqueId than their own state’s instance of the Ed-Fi API.
Currently, source systems must store a mapping of identification code(s) to the correct StudentUniqueId for a given contest.
Apparently, Texas has standardized on using the state ID
And in California they always use the local ID
Course ID might also be a problem in Texas
Can we get more information about who is having a problem with this?
EDITORIAL NOTE: after the meeting, we noticed that our two assessment provider TAG members were absent from the meeting. -SF
Michigan with NWEA
Some states do not have a unique ID for staff
In general, assessment rostering is difficult
Two approaches were presented for helping the client applications deal with this:
Add ability to query Student, Staff, Contact, and EducationOrganization by identification code.
Cache mappings on the API side and automatically translate.
Like ODS/API version 2
Option 2 is attractive to most… however, what about uniqueness?
Two different source systems can, today, insert the same identification code for a student.
This leads to possible cross-contamination of student data.
May need to make a breaking change to the Data Standard in the future to better support the multiplicity of student identifications with strong referential integrity and uniqueness. See DATASTD-1913.
Tentative conclusion:
There is a real preference for option 2, but the (a) enforces greater complexity in the Data Standard and the API application and (b) poses high risk in terms of data governance. Therefore, option 1 would be better at this time.
Lineage
We were unable to cover this topic; it will be postponed to the next meeting.