Entry tags:

fedora, olpc, sugar

La Dolce Vita: chapter 9, in which our hero thinks more carefully about our educational mission.
I just spent my weekend reading an exceptional book: Disrupting Class by Clayton Christensen, Michael Horn and Curtis Johnson. Anyone involved in OLPC or Sugar should take the time to read this book.

To summarize, as briefly as possible, the key arguments that I took away from it:

1. It's impossible to attack the education beast head-on. You can't go to the local school board and say "use computers the way I say please, thanks," because that approach never, ever works. Disruptive innovations are seldom developed to replace entrenched competitors; they first address underserved markets, where the status quo is essentially nothing. They prove themselves in this alternative market, grow over time, and ultimately they reach a tipping point at which they deliver functionality that is "good enough", ultimately pushing the incumbent aside. History is full of these examples: the transistor over the vacuum tube, the personal computer over the mini-computer, and so on.

2. There are plenty of underserved markets in American education. With the push towards raising standardized test scores -- arguably at the expense of actual learning -- these underserved market opportunities will increase; as schools are forced to spend more of their budgets on increasing basic scores in basic competencies, many of the "electives" will be pushed aside, and kids will have fewer options. Homeschoolers also represent a vast and growing untapped market.

3. Computer-based learning has the potential to fill these underserved market niches. Johnny wants to learn Arabic to help prepare for his future in the diplomatic corps, but he lives in the middle of Nebraska and there's no Arabic instructor for 500 miles? This is a dream scenario for computer-based learning.

4. Classroom-based teaching is, by its very nature, necessarily monolithic. Teachers can't afford to cater to individual students; textbook authors must write to an audience of hundreds of thousands to make writing a text financially worthwhile; school is broken into twelve grades, that a kid either passes or fails; predictable inputs and outputs; the whole system is built to run schools as though they were factories.

Computer-based learning, on the other hand, has the opportunity to be modular. Subjects could be broken up into small modules, each with its own assessment capabilities; therefore, student mastery of one module could be assured before the student would be allowed to move on to the next. Different modules could be constructed differently to accommodate different learning styles (visual, textual, auditory, etc.), so long as an equivalent assessment of skills mastery could be made. Online tutors could be made available to help with particular subjects, and these tutors could potentially scale much more effectively than teachers.

This is a poor and limited synopsis of a brilliant book. Go read it. It's thoughtful, groundbreaking stuff, and it has a lot of implications for the development of Sugar and other educational platforms. The most interesting implication, for me: perhaps we can stop assuming that we're developing Sugar only for kids in the developing world, and work instead on making Sugar valuable in our own communities.

Imagine, if you will, a Sugar educational wiki that has the following content:

+-------------------------------------------------------------------+
| 4th grade mathematics modules for Sugar!                          |
+---------------------------------+---------------------------------+
| Long division                   |   division-patrol.xo            |
|                                 |   divide-and-conquer.xo         |
|                                 |   divider-man.xo                |
|                                 |   mongo-teaches-division.xo     |
+---------------------------------+---------------------------------+
| Fractions                       |   fraction-patrol.xo            |
|                                 |   mongo-teaches-fractions.xo    |
|                                 |   denominator-dominator.xo      |
+---------------------------------+---------------------------------+
| Negative Numbers                |   mongos-number-line.xo         |
|                                 |   two-wrongs-make-a-right.xo    |
+---------------------------------+---------------------------------+

Now imagine that each of these activities has (a) a testing mechanism that allows demonstration of mastery; (b) a mechanism to record that data over time, for use by a teacher/parent charged with oversight of the child's progress (like, say, the Sugar Journal); and (c) a stamp of approval, provided by some lightweight/sensible approval process that allows teachers/parents to say "yes, this activity teaches the subject, if not perfectly, at least well enough."

Now imagine further that this wiki has the structure for a child's entire K-5 curriculum, across all subjects, with multiple activities to choose for each lesson.

Now imagine further that there exists a critical mass of these activities, written simply enough and documented well enough that a parent/teacher of reasonable intelligence and motivation could examine these activities, modify them to suit, or perhaps even create their own.

This is the kind of promise that Sugar holds, and we need to do a better job of articulating this promise, and figuring out the critical path that allows us to deliver on this promise.

I had a discussion with someone at FUDCon -- maybe it was David Farning, maybe someone else -- in which we discussed homeschoolers as a market to target with initial Sugar releases. That seems like a better idea every day to me.

Sugar folks, again: go read the book, please.