My Reflections on HEDW Conference 2019  – Shital Sharma, Clark University

First in a series from the recipients of the Cathy Lester grant, that funded their attendance at the  conference

I recently had the opportunity to attend the Higher Education Data Warehouse (HEDW) Conference in Michigan with the “Cathy Lester Attendance Grant” awarded to me by HEDW. Although I knew of the HEDW community, it was my first time attending the conference I was very pleasantly surprised to see how content rich it was. The conference addressed technical issues around warehousing and around machine learning, modelling, analytics, and visualizations using data warehouses. It also addressed issues around cultural and ethical challenges on data governance arising out of the decision to warehouse. Some sessions also addressed the advances in using cloud for warehousing purposes and the challenges that come with such changes.

I found these sessions and the subsequent discussions extremely timely and helpful.  I am myself dealing with a lot of these issues at Clark as we transition to a data warehouse with automated and recurring analytics and reporting structures. I found the experience of colleagues from other schools very inspiring, especially on questions surrounding automation of reports, prediction exercises and around governance issues. Discussions on how schools, large and small are dealing with cultural changes around data governance, how they are dealing with the challenges of balancing ethical and privacy considerations against the benefits of a data driven or a data informed decision making culture around higher education outcomes like retention were all so relevant and worthwhile. Also a takeaway for me was how colleagues are dealing with challenges on data quality when data warehouses are already in place and stakeholders are consuming live data with potential issues.

From the HEDW conference, there were several specific takeaways for me based on Clark’s needs. I outline them below:

  • A key idea I saw successfully embraced, and one I hope to implement at Clark, is a decentralization of the governance process – especially in generating data definitions and ensuring data integrity. The setting of data standards and definitions for data coming from functional areas within the institution that generate the data made the process more successful. Users of the data often had inputs towards an agreement on data definitions, however the primary responsibility of safeguarding the integrity of the data including how it is defined stayed with the individuals who generated it.

Larger governance concerns such as defining student gender or defining the standards for data privacy were more successful with a collaborative democratic process through a governing body represented by all functional areas that generate data.

  • I found most peers and colleagues also struggling with the process of data democratization and there were valuable lessons I learnt from their experiences. Training and constant communication of advances in reporting seemed to be the best way forward. Also, understanding of audience needs and development of creative and intuitive visual reports helped move data informed decision making culture forward for many. Democratizing data without adequate governance seemed to be the biggest pitfall to guard against and the shared experiences of colleagues helped me to understand how progress here can be achieved.
  • I found sessions that discussed the use of technology and analytics/machine learning to drive strategy very inspiring. There were several examples of this in different sessions. Automated machine learning tools and continuous analytical reporting were being used to mitigate retention risks, or reporting on student outcomes. Partnerships between IR and IT here was the essential ingredient for success. A shared vision between IR and IT of information dissemination and consumption was a large contributor to success.

It was also inspiring to learn how institutions are taking some matters in their own hands. With open source platforms, schools are no longer waiting on large commercial vendors to make system-wide solution upgrades. Instead, schools are making targeted efforts to upgrade piecewise solutions with in-house solutions built on these open source platforms. This was especially relevant to my experience since Clark is a small private school and there is a need to be innovative to succeed in many of these areas.

Finally, I believe I made a great decision by joining the HEDW community. While the conference was a great experience for me, I am looking forward to the benefits of HEDW that extend beyond the conference to a great community of colleagues that I can turn to for inspiration, resources and advice when I need it.

 

Shital Sharma

Assistant Director, Strategic Analytics and Institutional Research

Clark University