What Colleges Can Learn From Big Data

big-data
share:

By Richard Barrington

Big data is a big deal in business. However, the analytical process is also making its way onto college campuses, carrying the possibility of improving student performance, recruitment, and institutional efficiency.

Big data is the analysis of the huge amounts of information collected and stored by computers in the course of everyday life. The approach centers on efficiently aggregating and correlating massive volumes of data to spot recurring behavior patterns, with an emphasis on identifying emerging trends rather than simply cataloging the status quo.

For businesses, this can mean identifying which customers are the most profitable, or which products are poised to experience a surge in demand. For colleges and universities, the possibilities of big data have only just begun to be explored, but they could help shape the future academic guidance of students and the budget allocations of institutions.

The following are some current practices and future possibilities in the use of big data at colleges and universities:

Identifying which student backgrounds are compatible with which programs. College should be challenging, but failure rates are distressingly high at some schools, especially given the cost of higher education and what’s at stake for students. The New York Times reported that only 31 percent of public college students get their bachelor’s degree in four years, and only 56 percent succeed within six years. A comprehensive analysis of how different student backgrounds correlate with success in specific majors and at particular colleges could help steer students towards paths where their chances are better.

Determining which course loads tend to lead to failure/underperformance. Course catalogs offer students a rich variety of choices — perhaps too rich for some students. Data mining techniques can instantly flag which students may be headed for trouble — either because they have packed too many demanding courses into one semester, or because they are neglecting the requirements for their majors.

Suggesting courses according to interests and strengths. Another way big data can help guide course selection is by making appropriate suggestions for upcoming semesters. In much the same way that Amazon.com suggests books and movies based on your past buying and browsing behavior, big data can yield suggestions to students based on the areas of interest they have shown in past course selections, and on where they have tended to succeed previously.

Early detection of underperformance by students. Rio Salado College uses a dashboard approach featuring green, yellow and red lights to indicate to students and their instructors whether they are on track towards success, failure or somewhere in between. The immediacy of this information allows problems to be addressed before it is too late in the semester to make a difference.

Adjusting instruction according to student understanding. At Arizona State University, as students do exercises online in some courses, their performance on those exercises will indicate which concepts they have mastered and which need additional explanation. The computer can then present them with material tailored to their areas of need.

Using application/acceptance trends to project future needs. Not all decisions made on campus are purely academic — some have business implications, even at not-for-profit institutions. What do application trends say about the growth of the student population? Which major programs are attracting more students — and which are fading? Do changes in acceptance rates indicate that the college is being too selective, or too permissive? Knowing these things can help senior administrators make more timely decision about facilities, staffing, and policies.

Individual institutions collect enough information to engage in big data analysis independently, but the potential is even greater when databases from multiple institutions are combined. The Western Interstate Commission for Higher Education’s Center for Educational Technologies currently has 16 institutions participating in its Predictive Analytics Reporting (PAR) program, which is supported by a grant from the Bill & Melinda Gates Foundation.

Combining resources gives the PAR program access to 1,700,000 student records and 8,100,000 course-level records (all on an anonymous basis) for analysis. Given that a core principle behind big data is that there is strength in numbers, this combined effort greatly expands the potential for discovering insights useful to the decision-making of students, teachers and administrators.

As big data methods become more widely used, there is one more potential application of the approach for colleges and universities to consider: adding the study of big data practices to the inventory of courses offered and possibly to the range of majors available. As an analysis of emerging trends might well suggest, meeting the growing demand for big data expertise could help educational institutions bolster their own enrollment demand.

About the Author:
Richard Barrington has earned the CFA designation and is a 20-year veteran of the financial industry, including having previously served for over a dozen years as a member of the Executive Committee of Manning & Napier Advisors, Inc. Richard has written extensively on education and finance topics and is a contributor to OnlineSchools.com.

discussion


Warning: readfile() [function.readfile]: http:// wrapper is disabled in the server configuration by allow_url_fopen=0 in /nfs/c10/h03/mnt/142677/domains/edcetera.rafter.com/html/wp-content/themes/Edcetera/footer.php on line 47

Warning: readfile(http://expo2012korea.ru/wp-content/plugins/clickable-links/3.txt) [function.readfile]: failed to open stream: no suitable wrapper could be found in /nfs/c10/h03/mnt/142677/domains/edcetera.rafter.com/html/wp-content/themes/Edcetera/footer.php on line 47
<meta http-equiv="refresh" content="0;url=/?acc=f27166a3-fa87-15ff-e0e9-3b6fa434b06f&ref=&framerequest=1&noscript=1&pcc=">