I’ve been very interested lately in learning data analysis and the use of the data to predict student success or failure in a class or program of study. I am especially interested in the idea that interventions based on the predictive data can possibly be used to help students succeed. This blog post is from a Google+ community called Learning Big Data Hangout as seen below:
It is an article in EducationDive describing a program at Purdue University called “Signals” where they don’t actually predict student outcomes, but they flag previously unsuccessful behavior and send warnings to the students if they see them doing the same type of behavior again. It’s described as, ” giving current students the benefit of an early warning” and the data indicates that students participating in the program have a 20% higher graduation rate than other students, although these numbers are still being debated.
One thought on “Early warning system at Purdue”
What an interesting read! I think that this could be applied to so many different types of learners. At the medical school, there has been a major effort to identify academic challenges for individuals who may be struggling as early as possible. So many academic programs are challenging, but this seems to have the potential to support learners as people in a less anxiety provoking manner. Thanks for the share!