Posts Tagged ‘MIT’

08
Jan

The Future of Education

Written by Rick Henrikson. Posted in Commentary, Productivity, Society, Technology

Robot TeacherThe most depressing class I ever took was freshman intro chemistry.  Granted, it was advanced intro chemistry (oxymoronic, but that’s how the course identifications work at MIT; extra numbers = harder, extra letters = easier).  So this was 5.112 (as opposed to the standard introductory 5.111).  So why was it so depressing?  I had learned the majority of this stuff already in my high school AP Chemistry course.  I had actually done rather well in the chemistry class, finding most of the material to be quite manageable and I scored as well as you can on the AP exam.  That’s obviously not the depressing part, though.

The sad part came when I realized I had forgotten a significant amount of the material I had mastered only a couple of years earlier.  Moreover, I was finding the material even harder the second time around.  This made me come to two harsh realizations:

  1. My high school instruction was better than the equivalent MIT instruction for this particular course.
  2. I can forget something pretty quickly, particularly if I’m not using it regularly.

This really marked a turning point in my education.  For the first time, I felt the frustration of viewing previously familiar material with virgin eyes.  Realizing how short-lived any particular piece of knowledge could be, I decided I would no longer sweat the small stuff.  I didn’t kill myself memorizing and practicing things that I didn’t find interesting or relevant for my near-term future.  I grabbed the big picture, and delved in deeper just where I felt like it.  As a result, MIT was a very pleasant experience for me.  Unlike my indiscriminately intense study habits in high school, I decided to focus on just the parts I cared about, knowing that I’d have to relearn anything that I needed to actually use in the real world.

This obviously is not efficient.  Both the impersonal method of instruction, and the (effectively nonexistent) means of knowledge retention in the current system leave quite a bit to be desired.  So what will education look like in the future?

Education will be Personal

This is what we were all told in third grade while taking those silly tests to determine our learning style.  I probably leaned more towards the visual/reading side of things, and secretly questioned the validity of “kinesthetic” learning, but it didn’t really matter anyway.  These tests never had a real impact on anyone’s academic pursuits.  Education is still done one batch at a time, with everyone receiving the same content.  At best, some instructors mix teaching styles with some overlap in order to bring as many students along as possible (link), perhaps incorporating hands-on experiments, group discussions, and visual effects into a standard lecture.  But this isn’t the ideal solution.  When I say personal, I mean really personal.

Every student brings two wild cards to the education table: their learning style, and their current knowledge.  Now those are some pretty huge frickin’ variables, and I’d argue the latter is most important.  Yet, students are all presented with the exact same material within a given batch, at best receiving some sort of “refresher” or “catch-up” material.  This can’t possibly fill all of the cracks.  So you end up with a good percentage of students trying to learn new material on a foundations that is full of gaps.  Not the best structural engineering approach.  Especially considering you can’t really teach someone anything unless they almost already know it (forgot who said this – anyone know?).

Learning is incremental and it’s nearly impossible to really grasp new material before fully understanding the concepts preceding it.  This is partially why I much prefer to read a whole textbook from cover to cover rather than receive whatever bits and pieces my teacher chooses are important enough to cram into an artificially-imposed academic calendar.  Courses need to be personal not just to the student, but also to the material.  Students should accomplish work at their own pace, with recommended windows for milestone completion to help motivate them along.  The key is to always be progressing and retaining what you are learning, not necessarily to move faster than everyone else.  Regular assessment of understanding would be integral, allowing students to review any modular subjects they might be lacking.

As an intermediate step, video lectures will become much more popular, and we’ll eventually see the “best” Physics 1 lectures rise to the top.  This is already happening, with recent studies showing that 82% of students at University of Wisconsin-Madison would rather watch video lectures, with 60% saying they would even be willing to pay for those lectures.  Their reasons generally were linked to a more personal experience (watching lectures “on-demand”, making up for missed lectures, etc.)  But eventually, the standard lecture format will have to give way to more interactive media that tests and reinforces throughout the teaching process.

Interactive Learning will Take Over

While working at Lawrence Livermore National Labs, I took a few online training courses that involved interactive material.  The interfaces definitely weren’t ideal, but they were a step in the right direction.  Users were presented with flash-based tutorials followed by simple quizzes to reinforce key topics, with some navigation controls to help with reviewing material that was not adequately retained.

Imagine how much further this could be pushed.  You could open up a video of a lecturer speaking, with interactive tutorial elements playing on the side.  These could be standard graphs, figures, and videos, or more complex boxes taking in user inputs to produce simple visualizations that explain a concept much better than the waving of a hand or the scratching of chalk.  MIT has some of this kind of content associated with their courses, but it’s definitely not as well-integrated into individual curricula as it should be.  It’s fairly clunky to have to go back and search through a list of visualizations when you’re first learning (or subsequently reviewing) a topic.

So now you have a student immersed in an interactive lesson, maybe even taking advantage of some new Minority Report-style interface tools being developed by a few companies.  Throughout the process, students can be prompted with questions to confirm they are grasping a concept before moving onto the next one.  In large lectures, this doesn’t happen.  If you get lost somewhere, you remain in the dark for the rest of your miserable time there.  You could ask a question, but that’s a fairly inefficient solution in a large room of students where many people are not lost.  But with gradual questions integrated throughout the process, it’s easy to identify any stumbling points.  The software could even be smart enough to break a question down into component concepts, asking a second series of questions, and a third, and a fourth, and so on, until the root problem area is identified.  The student can then review that area until he or she is ready to return to the work at hand.  That’s real no child left behind.

Obviously there are some subjects that are more readily amenable to this new education platform.  Mathematics, language, and the sciences are all excellent candidates.  Some components of humanities education could also be addressed, with some modification.  Writing might use peer-based editing and assessment (much like many writer groups that are being formed online today).  Artistic and physical instruction can also be addressed with a range of new input devices.  The Wii is great for a lot of things, and there’s finally an educational game for guitar hero with an actual guitar.  These kinds of devices could eventually be integrated into a complete, interactive learning environment that is much more personalized than anything that could be offered in batch classroom settings.

Optimized Review will be Critical

A number of studies have emphasized the value of spaced repetition for memory retention.  The basic idea is that, after you’ve learned something it is very easy to remember upon review the next day.  Then, as time goes on, it becomes harder and harder to remember until you have no clue what it was anymore.  It turns out it’s probably optimal to review this material right before you’re about to forget it.  With spaced repetition, you review the concept at optimally-designed intervals to make sure you never forget the concept, with minimal time expenditure.  A number of companies have been developing software to help push spaced repetition, but it has mostly been limited to desktop applications with flashcard-style learning (great for language, but sub-optimal for most other things).  Smart.fm has received quite a bit of press lately for applying these strategies in a simple, web-based platform.

The successful integration of optimized review right into the learning process will have a huge impact on education, helping students to actually remember most of what they’ve been taught, and saving all of the wasted time catching everyone up at the beginning of every new semester.

It will be Cheap and Ubiquitous

Right now I’ve got a phone in my pocket that is more powerful than the average computer was a few years ago.  And on that phone I have access to millions of bits of absolutely free information, anytime and (almost) anywhere.

Of course many people rely on Wikipedia as the trusted source for a first go, but expert-produced content is making its way into the free space.  There are a number of sources for free video lectures, including my personal favorite.  An MIT alumnus has even started a Youtube channel that offers comprehensive instructional videos on everything from chemistry to differential equations to banking (1000+ videos!, thanks for pointing it out, Jamie).  And recently, open source textbooks have gained some momentum with Flat World Knowledge and Wikibooks.  Our very own Governator even pushed an initiative this year to get open source textbooks in high school classrooms throughout California, with the hope of ensuring high-quality and affordable education for everyone.

Ultimately, I believe educational content will become extremely cheap or free.  But I don’t just mean video lectures and textbooks.  I’m talking about entire educational programs with web-based content that’s optimized to improve students’ learning.  Initial investments developing these programs will pay off as they drastically reduce many other economic burdens imposed by the traditional education industry.

And being web-based, these tools will be available to everyone with an internet connection, a number that is continuing to grow.  Phones could even be used for rapid review of appropriate content.  You could run through some vocab or quick math problems while waiting in line.  And it could even become addictive if presented in a game format.  The popularity of educational games on devices such as the Nintendo DS demonstrates that people are both willing and eager to apply their brains to constructive problems in gaming environments.  So education will be cheap, everywhere, and addictive.

But Classrooms Still Have a Place

Moving to an entirely digital education would obviously have some terrible repercussions for social development.  From the beginning of my time at MIT, I realized the reason the place was special wasn’t because of any fancy machines or brilliant lectures.  It was special because of the connections you could make with some really amazing people.  They used to say that at MIT there are three things that take up your time: Sleep, Social, and Study.  You can only choose two.  Anyone in my freshman dorm can tell you that Social ranked pretty highly for me, with Sleep taking a bit of  a back seat.

Although some peer discussion could take place online, students will need actual human interaction to prepare them for the inherently collaborative nature of modern working environments.  That’s why these tools would largely have to be an enabling supplement for higher-level discussions and projects in a classroom setting.  Students could complete 80% of the learning on their own, with teachers and parents monitoring their progress via web interfaces.  Then they could go to class knowing they have something substantial to contribute to bigger, more integrative goals.

So when do I get my robo teacher?

I’ve primarily been discussing ideal education systems for the future, but what’s practical in my lifetime?  The biggest issues may lie in the fact that education is a huge industry.  And like any huge industry, it has a lot of inertia that will take time to adjust.  AcademHack has a great video presenting the issues our outdated “knowledge creation and dissemination” system will face in a modern, connected world.  There are going to be some tough growing pains, much like we’ve seen with the recording industry, the film industry, television networks, and publishers.  But clearly a lot of changes have to be made to reach an optimal system.

These tools will likely be implemented on a more individual basis in the near-future, as supplements to traditional schooling.  However, it’s clear that our nation is moving towards efficiency by technology and personalization.  We’ve been trying to get away from batchucation (coining a term – education in batches), and we finally have the web-based tools to make it happen.  I’m excited to start using some of them to finally refresh all of the material I’ve inevitably forgotten.

Reblog this post [with Zemanta]

Tags: , , , , ,