Post date: Jan 3, 2016 3:35:22 AM
Blog Post for January 2nd, 2016
As I prepare to head back to school, I’ve been thinking about all of the student data that gets generated from by online learning systems. Maybe you’ve heard the buzz phrase “big data” but few people seem to grasp what this means; however, for those of us in Higher Education, using sophisticated Course management Systems like Blackboard, Angel, or Canvas, big data is something we deal with all of the time.
As I wrapped up the fall semester, I had half a dozen students who were struggling and opted to do an Incomplete: this gives them a few weeks into the next semester to complete assignments. So even over break, I was checking email and logging onto Canvas, our CMS, to see if any of my students had submitted work, and fortunately, several had!
When I log onto my course websites, there is an innocent looking category called People on the navigation bar that lists the students and the teacher: the students see only a list of names, while the teachers have access to a great deal of data about each student. Among other things, I can see the following information once I click on a specific student’s name:
Furthermore, once I'm looking at the data about one students, and click on the link for analytics, I can see an overall pattern in communication, submissions, and activity: a bar chart that shows the days the student has logged on and worked.
Some of this information is extremely useful. Some of it is mildly creepy. When I mentioned it to some of my Education students, they were intrigued and a few seemed disappointed that they could not see this information. A few were somewhat unnerved and not sure what to think about it all.
Is there potential to understand student behavior based on this data? Certainly: when are students more likely to work on their online courses? I have learned to stagger due dates: assignments are not all due on Fridays, because some students work all weekend. I have students working third shift, while others are stay at home parents who do not work on classes until children are napping or down for the night. I have students who live out of state—and in different time zones. I have had military wives and soldiers enroll in my online courses. I have also had many students from other colleges take my courses online, for a variety of reasons. Looking at the data from all of these students has helped me to make necessary adjustments along the way.
However, this data does not address some key issues--
Unfortunately, there is no data generated by Canvas to address those issues, unless we survey the students who fail. However, I worry about the way that for-profit companies may use student data. As it turns out, I’m not the only one. Michael Nanfito wrote a post on LinkedIn, “Ethical Implications of Big Data and Learning Analytics in Higher Education.” He cites several examples of the big dollars being spent to acquire tools to gather and analyze student data.
He also cites Willis, Campbell, and Pistilli , who developed “a set of six questions that help inform the consideration of implementing of big data in campus learning systems:
I enjoyed reading Nanfito's post and will probably point it out to a few people at my college. I do believe that students should be better informed about the data they are generating every day as they participate in their courses online, and that teachers should get more training to understand how to best use the resulting data. In addition, we need answers to the questions I posed earlier regarding the students who fail courses by their lack of participation: in spite of all of the information that I can review about my students, it is their silence, their lack of participation, and their absence that most frustrates me.
I have been a technology enthusiast for my entire teaching career: while the technology has evolved from marker boards and overhead projectors to computer classrooms, laptop carts, iPad carts, and the use of Online tools like Blackboard, Angel and Canvas, technology is the tool. It is the human connection between teacher and student and the course content that engages students. The most significant data point for me has been to watch students in the first three or four weeks of classes: I can predict who will succeed and who will drop largely by whether or not they come to class and turn in assignments. Imagine that: companies are investing millions of dollars in learning analytics, which certainly provide useful tools. However, they might also learn a great deal by simply chatting with some veteran teachers like myself and some of my friends, who have piloted a number of technology projects over the years.
Data must be used carefully and ethically: students need to be informed about the way it is being gathered and analysed and colleges and vendors alike face many challenges in making sure that privacy is guarded and students are not exploited. In addition, I hope more attention will be paid to the puzzle of those students who fail to participate in these online systems, and in doing so, fail their courses.
For more information,
Nanfito, Michael. “Ethical Implications of Big Data and Learning Analytics in Higher Education.”
Oct 2, 2015. LinkedIn.com
Last updated January 2, 2016