OWD 2012- 4 -Four years of learning analytics in practice- John Campbell
Download as PPTX, PDF1 like387 views
This document summarizes John Campbell's presentation on Purdue University's use of analytics to help identify students at risk of dropping out. It discusses how Purdue developed a tool called Signals to detect early warning signs in student performance data and provide targeted interventions. Analysis of student cohorts found that those who received Signals interventions had significantly higher retention and graduation rates compared to those who did not. The presentation provides advice on how other institutions can build similar analytics capabilities and highlights both the opportunities and challenges to developing actionable intelligence from student data.
OWD 2012- 4 -Four years of learning analytics in practice- John Campbell
- 1. Analytics in Practice: An Institutional Perspective John P. Campbell Associate Vice President – Academic Technologies Purdue University
- 2. About Purdue
- 4. Data rich, information poor
- 5. Challenge: How do you find the student at risk? http://www.youthareawesome.com/wp-content/uploads/2010/10/wheres-waldo1.jpg
- 6. Challenge: How do you find the student at risk? http://www.youthareawesome.com/wp-content/uploads/2010/10/wheres-waldo1.jpg
- 7. Our Focus Signals detects early warning signs and provides interventions for students who may not be performing to the best of their abilities before they reach a critical point in their class.
- 8. Signals Approach Student Preparation + Student Engagement/Effort + Student Performance Messages + interventions
- 9. http://bit.ly/signalsdemo QUICK VIDEO ON WHAT COURSE SIGNALS IS & HOW IT WORKS
- 10. Signals after 5 Years Fall 2007 Cohort Retention Rate Number of Cohort Average SAT Signals 1 Year 2 Year 3 Year 4 Year 5 Year Size Score Courses No Courses 5,071 83.34% 73.04% 70.34% 68.92% 68.74% 1155 At least 1 1,629 96.44% 93.80% 89.93% 88.03% 86.80% 1130 1 instance 1,377 96.44% 93.32% 89.18% 87.00% 86.49% 1134 2 or more 252 96.43% 96.43% 94.05% 93.65% 88.49% 1103
- 11. Four & Five Year Graduation Rates Fall 2007 Cohort Graduation Rate Number of Signals Cohort Average 4 Year 5 Year Courses Size SAT Score No Courses 5,071 42.06% 61.39% 1155 At Least 1 Course 1,629 43.52% 74.34% 1130
- 13. http://bit.ly/SignalsFaculty QUICK VIDEO ON FACULTY PERCEPTIONS OF COURSE SIGNALS
- 14. Actionable by the student
- 15. Student Perceptions • 89% of students that experienced Signals had a positive experience • 86% of students said that the benefits outweigh the drawbacks in Signals • 73% said they would like Signals in every course • 61% of students felt they got a higher grade in a course because of Signals • 58% of students said they sought more help because of the personalized interventions • 74% of students believed that their motivation was directly affected by Signals
- 16. Student perceptions • “It was an interesting concept to use a stoplight to let me know how my progress was going. Just the simple color change was enough for me to know how I was doing in the class.” • “Seeing green motivated me to continue to maintain my positive progress in the class, and once I saw yellow, I knew I had to increase my efforts to do better in the class.”
- 17. Student Perceptions • “It was nice knowing my overall grade and progress in the class at the beginning, middle, and end of the course and allowed me to step up my effort before it was too late to help my grade significantly” • “It made me feel like the instructor was aware of me and was interested in my progress.” • “Signals helped me become more motivated towards the class and I raised my grade 20 points!”
- 18. How does a campus build capacity?
- 21. Institutional Challenge: Focus on Actionable Intelligence
- 24. New Possibilities • Using data that exists on campus • Taking advantages of existing programs • Bringing a “complete picture” beyond academics • Focusing on the “Action” in “Actionable Intelligence”
- 25. How to Begin – My Advice • Actionable intelligence • Moving research to practice • Fostering change in institutional culture • Understanding the limitations and risks • Take a long term view • Be part of a community
- 26. How to Begin – My Advice • Keep it simple • Start with some manual steps before developing/purchasing the “ultimate” system • Have a diverse team • Don’t forget the faculty and students as part of the process
- 27. SoLAR Annual Conference • 8-12 April, 2013 • University of Leuven in Leuven, Belgium.
- 28. Final Thought: Obligation of Knowing
- 29. Thank You John P. Campbell Associate Vice President – Academic Technologies Purdue University jpcampbe@purdue.edu http://www.itap.purdue.edu/learning/tools/signals/
Editor's Notes
- #8: It also reinforces strong students by telling them they’re doing well and should continue working in the same manner in the future.The system can be run as soon as there is one grade in the grade book. Many instructors want to provide feedback as to where students stand after the first test. Course Signals allows for students to know where they stand in a class in the weeks leading up to the test and the weeks after the test so that they can continue to make appropriate changes (more studying, visiting SI sessions or office hours, coming to class…) in an effort to improve the grade. The more feedback students get, the more likely they are to take advantage of the resources provided and, in turn, the more likely they are to earn a better (or simply passing) grade.
- #9: Magical SSA ;-) I will animate this as well