Guest Post: Big Data in Higher Education

Today’s guest post is from Brian Parish, President of IData Inc. His blog post shares a realistic view of the opportunities and barriers to the Big Data movement in education. 

As the owner and president of IData, Inc., Brian Parish is a well-respected thought leader in the Information Technology (IT) and Institutional Research (IR) industries for higher education. Parish combines his technical experience, innovative creativity, professional network and management skills with the goal to build the industry’s leading consulting firm for higher education. After launching IData in 2004, Parish has worked directly with nearly 200 colleges and universities to increase their productivity with creative and innovative custom technology solutions.

Higher education institutions aren’t homogenous entities. Most people think of prestigious private institutions or large state schools. Those only represent a small percentage of the country’s institutions. There are more than 4,500 accredited institutions in the United States and around 20M students enrolled at any time. Nearly half of those institutions have less than 1,000 students. There are 4-years, community colleges, state schools, privates, etc. Most of the smaller schools have limited IT resources and enormous demands on their time. Over the past decade, schools have focused on gradually transitioning from simply looking at their data as “academic records” to now looking at a complete enterprise-wide data management plan. This transition hasn’t been easy or universally successful. However, there are pockets of institutional success and a wide-range of vendors entering the market to help schools. It is important to understand this picture before talking broadly about the role of Big Data in higher education.

Schools now possess “Big Data.” The challenge is to find a path to use it. More of the student, faculty, and staff interactions within institutions are being captured in data systems. Most schools ran on a single student information system that managed registration, billing, transcripts, etc. From recruiting to roommate matching algorithms, specialty software applications have been coming on-line to cover every aspect of the student experience. These systems offer data sources for a complete picture of their students. There is, however, limited data integration and span of control between these systems to allow for truly effective analysis.

The promise of Big Data in higher education:

  • Retention:  More data on students will let schools develop more accurate models for predicting success or risk for student retention. Having more up-to date classroom and activity data will help to identify at-risk students earlier.
  • Recruiting:  The ability to use more sophisticated algorithms and data analysis to help target the right applicants can help reduce cost to a school, raise the quality of applicants, and clear the mailboxes of unwanted college recruiting packets.
  • Course demand:  Predictive analytics and richer data can help a school predict future course demand.  This allows schools the opportunity to ensure students get all the courses they need to graduate on time and also reduce costs for low-enrollment courses.

Higher education’s barriers to success with Big Data:

  • Silos:  Many of the new data systems that are being used at colleges are Software as a Service, SaaS, applications that aren’t inherently connecting to the main enterprise systems, which can often lead to both technology and political silos that prevent combined analysis.
  • FERPA:  The 1974 Family Educational Rights and Privacy Act has had the unintended consequence of standing in the way of serious research and innovation in higher education data analysis. Schools, terrified by liability concerns, have implemented strict rules about sharing and accessing educational data.
  • Trust & Understanding:  People discuss big data like it is simpler than “small data,” but it is not.  One of the big challenges is clear and agreed upon data definitions. If you can’t define the data and metrics well, then you cannot use it. Clear and transparent definitions will lead to trust, understanding, and use.
  • Campus Culture:  Schools are innovative places, but do not tend to move quickly at the institutional level. The culture of academic freedom at institutions often flows into the entire enterprise, and institutions find it difficult to implement top-down technology or data initiatives.

Who is doing it right?

  • Predictive Analytics in Retention (PAR) Framework:  PAR has been working to aggregate data from multiple institutions to develop metrics for predicting retention risks.
  • Georgetown University:  GU has one of the best 4-year graduation rates in the country. One part of this is the way they manage course demand through an innovative class bidding system that uses algorithms to award courses based on both need and desire.
  • Building trust and standardizing data definitions:  A plug for – IData’s own Data management tool for higher education.

Final Thoughts:

Big Data presents an enormous opportunity for institutes of higher education, but barriers stand in the way for many of those schools to effectively use it.  Success will come from the extremes: pockets of innovation from small departments operating on limited sets of new data, and large consortiums or government groups building metrics on aggregate data sets. Positive outcomes from these extremes will lead to adoption at the institutional level.