During the past several months there have been low background rumblings in the land of education and training, and the rumbling is starting to get louder. That is the sound of the learning world discovering what Internet professionals working in other vertical markets have known for years: The digital “breadcrumbs” that learners leave behind about their viewing, reading, engagement and assessment behaviors, interests and preferences provide massive amounts of data that can be mined to better personalize online experiences.
Learning analytics are all about applying statistics and research methods in the service of better informed, more accountable decision-making. We're talking about using descriptive statistics, inferential statistics and predictive statistics on the data that we collect about, well, just about everything associated with Online Learning. eLearning. Distance Learning. Whatever you call it.
While predictive analytics are commonly used in consumer settings (e.g. to help generate recommendations for Amazon shoppers and select movies on Netflix), predictive analytics are not yet broadly used in educational settings for assisting with activities such as course selection, or for predicting points in a student’s academic experience where they may be at a point of increased risk. We're not even very good at keeping track of descriptive information in education, a la Google Analytics. at least not yet.
But we're getting there.
Here are five things I know for sure about the rising tide of learning analytics and what this is going to mean for helping people make better, more informed decisions about learning. Yours, mine, theirs, ours.
A vast number of today’s online learning transactions are captured in Content Management Systems, Learning Management Systems, Student Information Systems and Enterprise Resource Planning systems. We know A LOT about what happens when students - and faculty - are online. Given what we know about web analytics in Google, and personalized recommendation engines in Netflix and Amazon, and the power of hastags in Twitter, "thumbs up" in Pandora and Facebook...why WOULDN'T we want to personalize our learning experiences in the same way?
I know for some that the most fun is going to be getting all crazy with the complex statistical modeling that this kind of work is going to require AND enable. But for a lot of us, what we are really going to care about is that learning analytics give ME a foundation for my own learning progress and completion. Set up the right way, these personalized predictively generated experiences will help ME stay on top of what I need to know. And THAT is very cool.
This isn't so much about collecting more information from people. It's more about getting smarter and putting the data we already collect to better use.
This is the hard part. This is the "admitting we may have a problem" part. Because once you take a look at what is going on, you need to be prepared for the fact that you are going to need to respond. To SOMETHING, anything. Ignore at your peril.
Same basic idea, just a different way of saying it. I will tell you all that I've been watching people squirming about how to report results in ways that show the not-as-good as well as the good in a couple of projects recently - enough to tell me that the benefits of learning analytics are going to be balanced with some big cultural barriers that will require direct and focused attention.
Yup. These data we leave everywhere we go...they really are going to change everything. Because once we KNOW what we know, there will be no going back.
SCORM and AICC have oddkes of features for tracking learner onteraction, but for most of us, accessing abnd using that data is some sort of voodoo magic that we Just Don't Touch.
I'm pleased to see tools emerging that will help us understand learner activity and engagement to better serve their needs.
Posted by: Steve Howard | April 23, 2011 at 06:16 AM