Data-informed decision making is a hot topic across all EAB memberships. Provosts wonder how to better allocate academic resources across campus, vice presidents of advancement strive to find a model that pinpoints individuals most likely to give major gifts, and facilities leaders want to improve classroom and laboratory space usage. Meanwhile, everyone’s trying to figure out what predicts student success to better identify and intervene with at-risk students.
Campus members often assume that the data to make those decisions is IT’s domain, and turn to CIOs to provide the data. However, collaboration between IT and other campus members is critical to improve the potential of data-informed decision making. Here are three reasons why.
1. Data that exists in distributed systems, such as Excel spreadsheets on a campus member’s desktop, fails to improve cross-campus decision making.
In conversations with campus leaders, they often cite data “silos” as one of the most challenging roadblocks to analytics success. Part of this comes from the old adage that knowledge is power—and having control over the interpretation of local data often provides functional units with that power.
However, the inability of other campus members to access that data for interesting analyses (e.g., are individuals who purchase tickets to student performances more likely to give to the university’s annual fund?) impedes analytics. Higher education leaders must communicate that data is a strategic asset of the institution, not of individuals or departments, and data should be responsibly shared rather than guarded.
2. Data is only useful to campus if it is accessible by members who might benefit from data analysis.
Higher education institutions collect vast amounts of data—from student grades to swipe card transactions. To promote data security, many campuses create a data gatekeeper approach to accessing data—campus members may only achieve access through a request process. Unfortunately, this process can sometimes take several weeks and campus members might not even remember why they requested the data in the first place. In an effort to balance security with openness, some progressive institutions have created role-based data access models to speed up access privileges.
3. The two moments when data quality can be affected most—when data is created and when data is used—are typically outside of IT’s responsibility.
Functional leaders need to establish standard data entry processes to improve data quality from the start. Further, institutions should create mechanisms to identify potential data quality errors, communicate these potential errors to functional units, and hold units responsible for correcting errors in the source system.