Data Is Not the Goal... & Other Reflections from HEDW 2018

    

Last week I had the opportunity to attend the Higher Education Data Warehouse (HEDW) Conference at Oregon State University with my  colleague, Joe Bertram. The conference brought together over 350 attendees from over 150 institutions. Whereas last year’s theme was “data governance,” this year’s was around “Artificial Intelligence (AI) enabled analytics” and “machine learning.”  I wanted to spend some time discovering and learning how colleges and universities are tackling the data warehousing and data analytics initiatives at their respective institutions.   

 To open the conference, the attendees enjoyed a very thought provoking keynote speech from Stephen Few, best known for his expertise on data visualization.  The keynote, entitled “Data is Not the Solution,” I found to be very apropos.  He argues that we continue to emphasize data as the solution, rather than people using data as the solution. Our ability to collect data has far outpaced our ability to understand data, making most data just noise.  Data is not the goal; being able to use data to make informed decisions is the goal.  You have to start with good relevant data but the resulting good decisions come from a process he calls “Data Sensemaking.”   

He left us with a few words of caution and wisdom which I thought were extremely relevant:

1)   When we surrender the ability to think to computers, we soon lose the ability to think for ourselves. Don’t get to the point where you are acting blindly based on the output of a computer. 
2)   Beware of metrics that resolve all complexity! Something’s up. The system is being gamed.
3)   Take time to Pause… Think…

 The conference also provided a wide variety of breakout sessions for attendees to learn about specific use cases. My HelioCampus colleague Joe  had the honor of moderating a panel discussion entitled “Redefining Data Warehouse and Analytics Operations in a New Era.”  Panelists included Dr. Yuko Mulugetta, Chief Analytics Officer at Ithaca College, Dr. Craig Rudick, Executive Director and Lead Data Scientist at University of Kentucky, and Kevin Joseph, Director of Business Intelligence at UMBC.  The audience especially appreciated Kevin’s introduction of UMBC as a basketball school (certainly due to their recent March Madness fame.) Some expected a retort from Dr. Rudick, but none was given, so the discussion moved to data.

As all three schools have become more tuition dependent in recent years, each began by addressing some of the pressing questions they are trying to answer for their institutions. For Ithaca, understanding how to use curricular, co-curricular and student life activities to better understand and articulate the “value” they’re providing to their students is a critical one.  Kentucky wanted to understand how they can better utilize merit aid to target and recruit those students who will ultimately graduate, rather than as a carrot for national merit scholars who will most likely transfer out. And, for UMBC, how do they use data to better understand what’s driving persistence and retention?  All scenarios present familiar challenges to most colleges and universities.

 Each panelist further commented on the need for both IR and IT to work together to solve these problems.  Relationships between IT and IR run the gamut; some relationships start out adversarial, some start out lukewarm, and some are well matched from the start, but efforts must be made to come together with a common vision if shared initiatives are to succeed.  

 The panelists concluded by discussing anticipated shifts in the makeup of analytics and BI teams over the upcoming years.  The consensus was there would be less of a need for “hardcore” SQL developers and more of a need for “hardcore, customer facing” SQL developers. This expanded role needs skilled individuals who, rather than responding to data requests with reports, instead ask questions like, “what are you trying to accomplish?” Dr. Rudick explained that for his team, “business analyst is no longer a single role, but now a fundamental part of everyone’s role.”

My closing assessment of the conference is that regardless of where each institution fell on the analytics continuum, it was an excellent opportunity for attendees to share their challenges, learn from peers, and hear more about options for solving their data warehousing and analytics needs.  Kudos to HEDW. 

Comments

Subscribe to Our Blog