The other day, I was building a Flash page for one of the projects I’m working on. The page was a game-type interaction — a cross between Wheel of Fortune and Hangman — in which the learner had to guess the letters which make up a term used in the course. The hidden term was something like “Corporate Controller” and the clue was something like, “Chief accounting position in the organization.” This interaction was mandated by our contract with the client, which usualy specifies a certain number of interactions per hour of instruction time.
As I was putting this together, I couldn’t help wondering how this was going to help the learner understand the purpose of the Enterprise Resource Planning software being implemented in their company. If it were me playing this game, I’d be thinking ETAOIN SHRDLU — the most common letters in the English language in descending order of frequency. I’d be able to fill in the “C” and the “P” when I’d exhausted the letters in the phrase and I’d move on to the next interaction with absolutely no retention of the point of that pointless exercise.
Will Thalheimer has an excellent article on this topic at e-learning magazine. After presenting plenty of research which shows what interactivity doesn’t do, he says,

So what is it about interactivity that makes a difference? The answer is that interactivity prompts learners to retrieve information from memory, and it’s this retrieval practice that prompts the learning improvements.

His article agrees with my observation about this interaction which the client demanded and we provided. He even goes so far as to say, “…questions about nonessential information actually hurt learning. They provide practice on retrieving the wrong information.”
Our product shipped last Friday. The client really like the “cool” interactions. Mea culpa.