Science company McGraw-Hill Education is collaborating, through it’s McGraw-Hill Education Learning Science Research Council, with Colorado State University (CSU).They are working on a research initiative to explore the use of advanced techniques in learning analytics and educational data mining to reduce the Drop-Fail-Withdraw (DFW) rates in science, technology, engineering and mathematics (STEM) gateway courses.

Unsuccessful course completion in these gateway courses is often associated with significantly lower retention and graduation rates. Capitalising on its research expertise, Colorado State University chose to work with McGraw-Hill Education to find new ways to use learning analytics to strengthen learning outcomes for its students taking these courses.

“Learning analytics is developing quickly as an area of academic research, and we want to use this type of research to solve strategic challenges at the university,” said Patrick Burns, CIO, Colorado State University. “By working with McGraw-Hill Education’s researchers, we hope to discover new techniques for solving the persistent challenge of high attrition rates in STEM gateway courses. We expect that the research will also benefit other courses and allow faculty to access data and insights in novel ways for enhancing teaching effectiveness.”

Dave Johnson, Director of Research and Analytics for Colorado State University Online, which has been leading CSU’s learning analytics research, commented, “Through our collaboration with McGraw-Hill Education, we are looking at how we can provide instructors with actionable, data-driven insights that will allow them to help students, at all levels, successfully complete their courses. We are already starting to see some exciting results and look forward to incorporating our findings in new practical applications.”

The research study, which is currently underway at Colorado State University, is focusing initially on testing and validating predictive models for course completion. The predictive model will be combined with a set of interactive insights for advanced diagnostics and intervention. These patterns can serve as early indicators as to which students will most likely complete a course and which ones are in danger of failing – and ultimately help instructors identify at-risk students that they can work with more closely to ensure course completion.