What is LARC?
LARC (Learning Analytics Data Architecture) is a research-focused data set containing information about students who have attended the University of Michigan since approximately 1996. The data is meant to help answer typical learning analytics questions on students, their academic careers, and their class outcomes. The data is divided into three areas:
- That which is constant throughout a student's academic career ("students" — e.g. gender, ethnicity, SAT test scores, high school GPA, and earned degrees)
- That which can change from term to term ("student terms" — e.g. academic level such as freshman vs. senior, academic career such as LSA vs. Law vs. Engineering, majors and minors, term GPA, enrolled credits)
- That which can change from class to class ("student classes" — e.g. subject, catalog number, credit vs. audit, earned grade, instructor, meeting location).
A complete data dictionary is linked from the Teaching & Learning data sets page. LARC data is delivered in U-M's Oracle database (like the rest of the Enterprise Data Warehouse) and also as flat files in an U-M Box folder.
How do I get access to LARC?
LARC access can be requested through OARS (Online Access Request System). Requests will be reviewed by U-M data stewards, primarily the Registrar's Office. Acquiring approval is easiest if requesters enter a brief abstract of their intended work, in OARS' Additional Comments section, to help approvers understand how the data will be used. Requesters can ask for the following roles:
- LARC Student Info
- LARC Student Info with Identities
Requesters should obtain Institutional Review Board (IRB) approval and sign a Memorandum of Understanding (MOU) describing their specific research interests prior to submitting their request. When access is granted, requester will receive access to Oracle tables for direct database access and U-M Box folders for flat files.
If the user does not have a department affiliation (e.g. student) the Dept ID field must be filled in with the department making the request in order for the UL to approve it. This Dept ID field must be entered BEFORE the request is submitted. If this is not done, then an incident needs to be opened for ITS Access & Accounts to enter the deptid (see screen shots below)
What tools can I use with LARC?
Any tool capable of connecting to an Oracle database can connect to the LARC data set. Our recommended tools are Oracle SQL Developer (SQL client) and Tableau (data visualization), but we have connected many other tools including SQL Workbench/J (SQL client), Denodo (data virtualization), and R (statistical computing). If using LARC flat files downloaded from the U-M Box folder, any software capable of working with large flat files should work.
How do I connect to the LARC database in Oracle?
Start by referring to communication you received when you first were granted access. That should contain connection strings and ID/password information to help you make a connection. If using a tool that supports TNSnames.ora files provided by MiWorkspace, edwprod.world is the correct selection. If you still need help for other connections, please contact the ITS Service Center. They have access to a Knowledge Base Article that contains detailed connection string information that we prefer not to share in a public document. If you prefer not to use a database connection, download the flat files from U-M Box.
Can you help me connect to Oracle? I'm having problems and can't make the connection.
Contact the ITS Service Center. They have support documents on connecting to Oracle, including LARC and other database assets. If you prefer not to use a database connection, download the flat files from U-M Box.
How do I find LARC flat files in U-M Box?
LARC folders should appear at the root level in approved users' U-M Box directories.
What about the Business Objects universe?
There is no BO universe for LARC. LARC was designed primarily with research users in mind and researchers want direct access to data without an intermediary reporting tool.
How often is LARC refreshed?
LARC is refreshed four times per year—approximately three weeks after the end of each academic term. The specific refresh schedule is available online.
Can I join LARC data with other data sources?
LARC can be joined with other data held in the data warehouse or other local data by linking through students' UMID or uniqname. LARC can be requested with or without identities. When requesting LARC with identities, UMID and uniqname will be included.
How do I interpret the data I see in LARC?
Refer to the data dictionary (linked from the Teaching & Learning Data Sets page) for technical metadata. Refer to the data glossary (coming soon) for user-contributed information. For interpretation questions, submit an ITS Service Center request to the LARC Dataset Support team and the question will be routed to the university's subject matter expert on your specific question.
How can I suggest changes/enhancements to this dataset?
Change requests for the LARC data set should be emailed to email@example.com for review by the data set's oversight committee.