Kathy Sierra posted earlier this month on How Much Control Should Users Have?

I can’t do better than to reproduce her graphic, which resonates strongly with me.

My own experience over three decades (secondary school teaching, graduate teaching, corporate training and development, independent instructional design, assorted performance improvement efforts) strongly suggests she’s right.

Not everyone wants to build an engine from scratch (or smelt the steel to shape the parts to build the engine); a lot of folks just want to drive the car. And even then, as when you rent a car, there’s only so much control you need to make you happy.

I’m not arguing for no control — Sierra does a good job discussing the tradeoffs. I’m saying that we need to balance the desired result for the user (or the learner) against the amount of choise we offer.

When I began with distance learning, designing mainframe-based training for Amtrak’s reservation system, I realized that a highly effective way to “individualize” the training was to build the content in logical chucks (“how to check if the train’s on time,” “how to calculate family plan fares”). If you don’t want or need to know on-time status, you individualize yourself right out of the course.

That turned out to be more effective and vastly more efficient than trying to create highly individualized paths within the courses. Instead, we focused our efforts on:

• Realistic simulations of the actual system
• A high degree of interaction, and
• Specific, detailed feedback when you got something wrong (or partly right)

Not that it’s an easy balance to maintain. Or, as pianist Artur Schnabel said,

The notes I handle no better than many pianists. But the pauses between the notes – ah, that is where the art resides!