UFIT Learning Analytics
Learning analytics is a relatively new field of study, with the result that the definition of “learning analytics” is still evolving; but Educause, the leading organization for higher education and technology, has defined learning analytics as, “the use of data, analysis, and predictive modeling to improve teaching and learning. Learning analytics examines data from various sources, looking for patterns and correlations that can provide insight to learners, instructors, and those who support them about how to improve learning. In this way, learning analytics can help students, faculty, and institutions achieve their respective goals.”1
The UF Learning Analytics Advisory Committee, a faculty committee which met from Spring 2015 through Fall 2017 considered a variety of definitions for learning analytics at UF, and preferred the definition, “the aggregation, distillation, and visualization of data associated with the educational process for the purpose of quantifying and improving learning, not just improving grades and course completion.”
Functionally, learning analytics involves collecting the data being gathered in various online teaching and learning services (Canvas, Turnitin, Voicethread, Pearson, McGraw Hill, etc.), aggregating that data alongside other data about students (e.g. Registrar and demographic data), cleaning and organizing the data to best support research, engaging in statistical analysis to determine patterns and correlations that can predict likely student success or lack thereof in specific courses, and delivering to instructors, students, and advisors timely information to enable intervention and changes to behavior when a student appears to be off-track for academic success in a course.
Why is this important?
There is a growing body of research that indicates that statistical models can, in some circumstances, predict likely student outcomes in specific courses; and that these predictions can reach levels of 75% to 85% accuracy as early as the third week in a term. Obviously, if a student is notified that they are off-track for success in the first few seeks of a term, that student will have more time to change how they are working and to seek additional help if needed. Likewise, instructors and/or academic advisors who receive such notifications have a greater opportunity to intervene with that student to assist and encourage more productive effort to achieve success.
Not only does such notification and intervention provide the opportunity for timely intervention and greater student success, such as higher grades and higher course completion rates, this can also decrease withdrawals and course repetition leading to lower 4 and 6-year graduation rates, improve student learning outcomes, and foster a more supportive learning experience leading to improved student retention and satisfaction.
Current Work: UF LEAD Dashboard
Current learning analytics work at UF is focused on providing more and better information about students and courses to the instructors teaching those courses. Future plans include proving information to students and to academic advisors.
The UFIT Learning Analytics Team has developed The UF LEAD dashboard for faculty. The first priority for UF LEAD is to replicate and improve, in both speed and accuracy, the data displays available in Canvas. We are also working on improving and expanding upon currently available information and functionality of the dashboard.
The UF LEAD Dashboard is currently available to all instructors using Canvas by logging into Canvas and in each course where you wish to use UF LEAD, click on:
Settings >> Navigation >> Click and hold UF LEAD >> Drag UF LEAD to the top navigation section >> Click Save
If you need assistance with this, you can call eLearning Support at 352.392.4357 #3 or email firstname.lastname@example.org.
In addition to UF LEAD, we have a statistician working on mining available data, “cleaning” that data, and running in-depth statistical analysis to identify patterns that may be of use to faculty in understanding their courses and their students, as well as correlations that may predict likely student grade outcomes in those courses.
The work of the UIFT Learning Analytics Team is regularly presented to a focus group of UF instructors who advise the Team and make suggestions about the dashboard, statistical research, and development priorities. This committee typically meets once per month to see demos of recent development, get updates on other work, provide feedback and recommendations to the Learning Analytics Team, and help set priorities for development.
Feedback is not limited to this committee: Any instructor can give us feedback and make requests or suggestions by emailing UF-LEAD-ATemail@example.com.
If you would be interested in serving on a Dashboard Focus Group, please send an email indicating your interest; and if you have questions or would like to discuss learning analytics at UF or in general, please feel free to contact Doug Johnson at 352.393.2308 or firstname.lastname@example.org.
- - - - - -
1 EDUCAUSE Learning Initiative, 7 Things You Should Know About Developments in Learning Analytics. Online at https://library.educause.edu/resources/2017/7/7-things-you-should-know-about-developments-in-learning-analytics.