Thoughts on Predictive Learning

When one of my piano students shows up for a lesson I am immediately interested in what they did during the week. More specifically, I’m interested in how they did what they did. I want to know how they practiced, not what they practiced – how much hands alone practice, how much repetition, did they use the metronome? count aloud? practice perform? There is one piece of data that I could take or leave – time at the piano. It’s just not that important to me. All my students ask me how long they should practice every day. My goal is to help them understand that this is the wrong question.

Elearning environments will soon consider how the learner has interacted with material. It will then predict what the learner will want to do next instead of simply keeping track of the boolean “has X been completed?”. It might suggest reading material based on the titles of previously-read material. It also might suggest a good short lesson plan based on previous session times. Predictive learning can react wisely when it asks, unemotionally, things like, “how long are you planning on staying?” and “would you like some reading suggestions?”

The human teacher has the advantage and disadvantage of being emotional. It’s obviously nice to connect with students on a human level, but emotions can cloud judgement of what should probably come next. I know I get tired of reinforcing habits in my piano students – so I skip something or omit telling them to keep their fingers in a certain position – even though they need to hear it again – because I’m human and forgetful sometimes. An LMS could potentially sense emotion – “John hates these articles because he spent only 7 seconds on them and his rate of reading is 80 words per minute when he’s interested.”

This is one reason I’m jazzed to see the Tin Can API solutions people come up with. What can we do with this new potential for information? The guys at SaltBox seem to be mildly obsessed with this and I can’t wait to see their solutions up close. Beyond a large-scale solution, I think there are lots of small applications this could be used for. If you have come up with something – I’m dying to see examples in action. I’ll be trying it out myself soon in little micro projects.

Related Articles:

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s