The Grass Can Be Green On Both Sides: Synergies Between In-House and Vendor Data Science Tools

Higher education institutions are often faced with the question of whether to use products from edtech vendors or to hire data scientists to develop tools internally. We will use a recent collaboration between the Enrollment Management and Analytics departments at the University of Arizona to highlight why this either/or framing can be a false dichotomy. We will begin by reviewing high-level findings from a comparison of two independently developed predictive models, one from a trusted long-term vendor and one developed internally by data scientists in UAIR. We will then focus on the learnings derived from this project. We will discuss topics including how this project facilitated better cross-departmental communication and understanding of the various predictive tools in place across campus and how they are used to support multiple missions; improving trust in both internal and vendor tools through reconciling them with each other; identifying key data points, variables, and trends that were highlighted by each approach; the relative strengths and weaknesses of each; and the positive synergies discovered by embracing both tools. We will conclude with an interactive discussion around vendor and internally developed products and how and why a strengths-based technology orientation can often lead to increased value for higher education institutions, particularly in the emerging areas of data science, advanced analytics, and artificial intelligence.