I just started reading Karl Kapp’s book Games, Gadgets, and Gizmos for Learning. Much of the book is about the differences in how different generations learn. Specifically, he compares the “boomers” to the “gamers.” One of the characteristics of “gamers” (called by others the net generation, or digital immigrants, or millenials, or probably some other names I’m forgetting) is that they learn by trial and error. They don’t want to be led by the hand through each step or to be told exactly what to do.
This is different from might be typical in previous generations. Natalie told me earlier today about how sometimes when she trained adults they were frustrated when she asked them to solve scenarios that weren’t exactly like their previous experiences. They didn’t want to apply existing skills in a new scenario, they just wanted to be told which button to push in what order. (Anything about generations and age is necessarily a generalization–I’m talking about overall trends. There’s lots of exceptions in any age group.)
I haven’t see any trends related to age in instructional design, but I wonder if there’s a similar structural difference between ID models. The traditional ADDIE model is a structured process; the steps mostly go in order. (OK, it isn’t completely linear; evaluation often happens at multiple points. But you get the idea.) Rapid development for e-learning is often about iterative prototypes. Michael Allen calls it “successive approximation” in his book. In other words, rapid e-learning is a method of trial and error. If a prototype doesn’t do what you want, just scrap it and do something else; it’s like hitting the reset button on a game.
These are definitely “thoughts under construction” here, but it seems like there is a pattern of moving from more rigid structures to more flexible ones. It’s more than just these two areas of course; it’s Britannica to Wikipedia, broadcast news to blogs, Dewey decimal to folksonomies. I haven’t seen anything that connects differences in ID models to generational differences though. Does anyone have anything about age and ID model preferences? Do you think successive approximation could be part of this broader trend towards the flexible and decentralized?