Interaction: not enough Cooks

In 1986, Donald A. Cook of Human Performance Associates gave a presentation at the Fourth Annual Computer-Based Training Conference in San Diego. I wasn’t there, and I never met Cook, but I’ve kept a third-generation photocopy of his presentation, Learning Principles and CBT Authoring.

For the rest of the post, I’ll let Cook speak for himself [though I may add clarification inside brackets]. You might find something of value yourself, and I may have enough to do more excerpts in another post.

Some lecture frames [non-interactive screens] are very ‘lectury.’ They crowd a lot of text onto the screen, they use condescending turns of phrase like “of course…”

Most important, the learner doesn’t have to do anything.

The field of CBT presents us with hour upon hour, or mile upon mile of “lecture” or “page-turning” or “text-y” screens…whose only incitement to student interaction is the repeated message: “Press return to continue,” or, sometimes, in a burst of libertarian spirit, “Press any key to continue.”

I have seen a self-styled tutorial accompanying a highly touted software product which could be traversed…by placing a book upon the space bar and walking away.


Authoring systems…will call such frames “information screens”…. Information is good stuff these days. Lectures, on the other hand…have fallen into disrepute. but they are the same thing…

The pattern of a string of lecture frames followed by a batch of questions is not very good. It would be better to begin with the questions… so that the learner’s first exposure to the important material comes after a look at the questions she’ll be held responsible for….

Test items, or as they are sometimes called, criterion frames… may be buried smoothly right in the middle of a teaching sequence…. an important guideline for the development of CBT: write all your criterion frames first, and put them in place — in the order which the learner will encounter them.

[In contrast,] a teaching frame presents a bit of information… and then asks for its application to a specific case…on the same screen. This…[has] the most powerful teaching potential in frame-based CBT….

I have often worked with instructional developers, slowly getting them to see the arbitrary and even crual aspects of telling the learning something on screen one and then testing for it on a later screen. Finally the magic moment comes when I say, “Why don’t you try putting both the information and the question about it on the same screen?”

There is a moment’s pause. Then the answer comes:

“But then, they’d get it right!”

Okay, this is Dave again. I’m back, because I want to summarize what Cook called a Tolman frame — a basic unit of interaction that includes:

  • Presenting information critical to some performance
  • Presenting an object for that information
  • Requesting a response

Here’s one example:

Gaul is divided into three parts, one of which the Belgae inhabit, the Aquitani another, and the third those whom we call Gauls but who in their own language are called Celts.

How many parts are there to Gaul?

You’ll notice that this is a bad example. The critical processing needed to answer correctly is about two points above plant life. It’s the instructional equivalent of “copy the word three.”

So here’s a better example:

The main tribes in Gaul are the Belgae, the Aquitani, and those whom we call Gauls but who in their own language are called Celts.

How many tribes are there in Gaul?

Notice the difference? The only way to answer correctly is to do something with the information about Gaul. You have to process it. Sure, counting the names of the tribes is a fairly simple process, but this is an example, not an online course about De Bello Gallico.

Desert-trek photo by ecreyes / Ernie R.