Big Data | Feature

5 Ways to Personalize the Student Experience With Big Data

Each time a student registers for a class, swipes an ID card, or visits with a campus advisor, he or she leaves behind a trail of data -- information that provides valuable insight into student habits, practices, and academic performance.

A growing number of institutions are now using this data -- and the analytical systems that make this information useful -- to boost the student experience, from improving assessment, to developing new instructional tools, to identifying students at risk of dropping out.

Here are five ways that two different colleges are currently using data analytics to personalize the student experience:

  1. Optimizing assessment data. Four years ago the Ohio State University College of Medicine got serious about using data to personalize the student experience. Before that, Eric Ermie, program manager for assessment evaluation for the Columbus-based institution, says students enrolled in courses, attended classes, took their final exams, and either passed or failed. If the latter occurred, a remediation exam was administered – again on a pass or fail basis. When the school switched over to an ExamSoft computerized testing system in 2009, it gained access to data that was previously untapped. "We knew a lot of the data was being left behind – including some points that we really thought could be useful," said Ermie. "When we gained the ability to categorize some of that data and leverage the metadata (i.e., the data about the data), we saw even more potential."
  2. Track and help repeat remediation candidates. To move beyond its traditional "pass-fail" exam process, Ohio State's College of Medicine started keeping tabs on students who repeatedly took remediation exams. To do that, the institution categorized its course-specific assessment questions and then related each query to specific knowledge, capabilities, and thought processes that it felt were required to answer those questions. Using its assessment system – which collects, analyzes, and then disseminates the useful information – the school developed and "tagged" the related learning areas that, in turn, allow professors to quickly see how individual test takers did on recall questions, which parts they are struggling in, and what additional support will be needed. "Faculty can see where the students are facing challenges," said Ermie, "and then help direct their studying."
  3. Allocate coaching support with limited resources. Judith Murray, campus executive officer at Altius Education and special assistant to the president at Toledo-based Ivy Bridge College, a 2-year, online associate degree program developed by Tiffin University and Altius, said the institution has been using data analytics for about a year. Working with a student population where 80 percent of enrollees are the first in their families to attend college, the school uses a "success coaching" approach that's focused on those pupils who most need the support. Coaches not only motivate students to complete their daily work, said Murray, but they also help pupils stay focused and motivated. In need of a better way to pinpoint the students who could benefit from success coaches, Ivy Bridge College has created prospective student profiles across three different risk categories. "With this information at our fingertips," said Murray, "we can put the coaching emphasis on students that need the most assistance and keep them on track. Over the last year we've also validated the accuracy of incoming student profiles and further refined the coaching process."
  4. Create longitudinal assessments. Ohio State's College of Medicine uses an integrated curriculum. Rather than teaching one course in anatomy and another in physiology, for example, faculty members focus their efforts around cardiology as a whole. "That's the best way to teach and learn medicine," said Ermie, "but it doesn't show us how a particular student is actually doing in anatomy." A final cardiology exam, for example, may only include 10 or so questions on anatomy. To fill in that gap, the school uses data analytics to look across 20 past exams (taken by the same student) to see if he or she is proficient in a specific subject area. "That gives us an accurate and stable picture," said Ermie, "and helps us better prep students for their US Medical Licensing Examination boards."
  5. Identify key knowledge deficits. When it started using data analytics to get more granular detail about student success, Ohio State's College of Medicine learned something interesting about test-takers: those that failed an exam the first time typically scored 90 percent or better on the second try. Some of that was because 70 percent of the test questions were identical to those on the first assessment. "They were only scoring about 66 percent on the brand-new questions, which comprised about 30 percent of the total test," said Ermie, who reviewed the data over time and realized that some students who advanced through the program were doing so with the same knowledge deficits they had when the failed the assessments the first time. To tackle the issue, faculty members took a larger role in helping to guide the study process and started using remediation tests that included 100 percent new questions. "We had a slight uptick in failure rates as a result," said Ermie, "but we've also found that those students who did pass the remediation exams wound up having fewer problems down the road."

About the Author

Bridget McCrea is a business and technology writer in Clearwater, FL. She can be reached at bridgetmc@earthlink.net.

comments powered by Disqus