Page Comparison

• Discuss data sharing agreement to support limited sharing of enrollment data between colleges

Caveat: Be careful to not treat MIS data like a transcript. May help identify if students are swirling.

Need to articulate the User Cases. Identify how students to potentially opt out. How colleges would collect and how it would get back to the tech center.

Erik happy to help craft use cases, get in front of legal counsel.

Denice uses the concept in the Strong Workforce, could be a use case.

Erik asking IT folks what use case should include; Chris Blackmore mentioned security would need to be included.

John has a PIC for student profile service. Has opt out and other services that could be used. Chris indicated there is a record on their student record, a flag in the SIS that says student does not want data shared. Chris’s SIS is Banner. Denise is homegrown moving to Banner. UDW program United Domestic Workers have students sign to share their data, as they register for classes, so the agreement can be universal or at the program level. Mark asking if it is typical or if colleges moreso using a universal opt-out.

Jason indicated that North Orange is Banner and has a field called Confidentiality Indicator, applies to FERPA block and all employees.

People are to collect their agreements and send to Mark.

Ting Layheng … informed consent.. students may expect their data would stay at the college.

Erik indicated a CCCApply field at end to share the data. Says CO is exempt.

Where students enrolled, not necessarily course or grade information, does that make a difference to the MOU or is it the same level of issue since it is solely registration data, curious. FERPA allows us to report attendance information. Not course level data. Could be under Audit and Evaluation category under FERPA which does allow exchange … student does have opt out ability.

Is it better have all the information or is it okay to only have a portion of data such as passed courses.

Jason posted APs regarding data privacy and security. Denise added Saddleback’s.

Need to address if the student approves at some colleges and not others.

FERPA… privacy block

Next steps: draft of bones from Erik at the next meeting in 2 weeksThe group discussed the feasibility of establishing a data sharing agreement to enable the data warehouse to track students who are enrolled at multiple colleges. This would provide insights into students' true course load and attendance patterns across institutions. Key considerations include:

• Defining the specific use cases and data elements that would be shared (e.g. just enrollment status vs. course and grade details)

• Ensuring compliance with FERPA regulations, including allowing students to opt out of data sharing

• Determining the logistics of implementing data sharing, such as handling colleges that don't participate and maintaining data security The group agreed to further develop a draft data sharing agreement for review at the next meeting.

• Review of new dataset to be released and supported use cases

Last meeting introduced Jack Thompson , presented a new data set he has a dataset to drill in on, discussion of use cases. Mark looking at use cases for colleges to verify the data is as it should look.

There is raw data, you would have to aggregate it, it does not come aggregated. Several Apply databases with unique attributes, application specification tables but then there is a separate model. such as a count of all applications. part 1.

Another point is time travel to support querying and “as of” date as of 3 months ago, 5 months ago, complete history stored.

Correllation between MIS and apply (already in replica but the replica is based on SSN) but has taken it further, and built a correlation table to use several methods to match students. Zip Codes, apply to college then attend, the correlation table will tell how it correlated the student to the student. a student to another student.

Students opting out was not considered. Jack mentioned he doesn’t have a way to view the data yet.

Biggest use case is correlation between MIS and apply. Able to tell what was the journey … who started and didn’t finish, started another and didn’t, did they attend more than one.

A data set within redshift with correlated model that goes to the other databases. Set of queries and views. Gender info, from all the apply databases, other attributes on the application.

Asked for use cases that would be useful. Denice does look at the funnel of application to completed, did they complete app to enrollment at the college.

Mark said we added a binary field of students started but did not finish. If applied at multiple institutions? Completed/didn’t completedeveloped that consolidates application data from multiple sources into a single model. Key features include:

• Correlating application data with MIS data to enable analysis of the student journey from application to enrollment

• Providing historical data with the ability to view data as of a specific point in time

• Offering a set of queries and views to help users understand the coverage and contents of the data set The group discussed potential use cases for this data, such as analyzing application funnels and understanding student mobility across colleges. The data set will be piloted with a subset of colleges before broader release.

• Pilot testing of new data set

Expectation once release other use cases would be identified but if this group could identify. Will deploy to pilot in next couple of months. testing would be minimal for queries, if data looks goodOnce deployed to production (early November), we would appreciate feedback. To Prod beginning of November. Pam willing to test.

• Confirm new meeting time starting with October meeting: 3rd Thursday of the month 1-2PM (based on results of Doodle poll)

AgreedThe group agreed to move the monthly meeting to the third Thursday of each month, starting in October. This change is intended to allow more participants to attend.

• Other updates

Agenda items for next meeting next meeting?

Questions and Feedback