The D37 Design Framework

The following framework was created by Chris Riesbeck. I have adapted it somewhat to reflect my own beliefs and experiences, but it remains very true to his original.

It is based on the belief that the design of technology to support learning is first and foremost about the design of activities. Once appropriate activities have been identified, the role of technology in supporting those activities and in encouraging learning can be considered.

The framework makes a few assumptions about learning.

Therefore, the design of learning consists of

  1. identifying skills and knowledge for which the audience has a legitimate need;
  2. identifying a task that the audience will recognize as useful or interesting and that requires them to have those skills or knowledge;
  3. designing an environment that will allow them to perform that task, experience failures if they lack the target skills or knowledge, and provide them with the opportunity to develop the skills or knowledge in response to their failures.

The framework consists of the following key questions that need to be answer when creating an initial design for a learning environment. They assume that you have already identified a task and an audience, such that the task is a useful one for that audience. The first three questions are about the task as it is conducted in the real world, that establish the challenge for designing learning activities. The second three questions are about the design of the learning activity.


Analyzing the task in the real world:

Designing a learning activity:



What mistakes do people make and why do they matter?

This is the critical question for understanding what people need to learn. If they already can do the task well, then there's no reason to teach them anything. If they can't do it, you need to zero in on what they have difficulty with.

What is the target audience doing wrong or not doing at all and why is this a problem? Examples of bad answers are "students convert fractions incorrectly" or "students think Vietnam is the capital of Japan." These answers don't say why these mistakes matter to either the student or anyone else.

In evaluating answers to this question, always plan to ask "so what?"


Why do they make these mistakes?

Given the performance failures, what causes them? Can anyone do this task well, or is this simply a hard task in a bad environment?

If it is a fixable performance problem, what's the cognitive cause? Possible answers include:

Explanations of causes of failure usually need to go several levels deep. Typical bad answers are either simply wrong, e.g., "people don't know driving while drunk is dangerous," or incomplete, e.g., "people don't know what good investments are."

In response to answers to this question, always plan to ask "why or "why not?"


Why don't they learn from these mistakes?

If people learn by doing and learn from failure, why aren't they learning from their failures here? What are the obstacles to learning in the real world?

Typical answers to this question include:

Bad answers to this question confuse causes of learning failure with causes of performance failure, e.g,. "they don't know careful planning ahead reduces execution time."

In response to answers to this question, always plan to ask "so why don't people learn to do better?"


What design can overcome the obstacles to learning in the real world?

This is where the process shifts from analysis to design. Given general characteristics of the performance goals, performance failures, and learning obstacles, it's possible and necessary to decide on appropriate activities and support structures.

For example, if people drive when drunk because they think their driving skills aren't that badly affected, you have to get them drunk and test their skills. If people invest poorly because of gullibility when listening to investment sales people, you have to expose them to many different scams, mixed in with reasonable offers, until they learn how to tell the difference.

Bad answers to this question apply inappropriate techniques. For example, driving scenarios emphasizing failures are exactly wrong for people who don't know to drive because they're afraid. Budget planning scenarios showing how rich you'll be when you retire are wrong for teenagers who could care less about retirement.

In response to answers to this question, always plan to ask "does this really address the obstacles?"

What makes this a meaningful task for the audience?

This question addresses the relevance of the task. The learners must feel that they are engaged in a meaningful task in order for them to engage in it sufficiently to learn from it.

In evaluating answers to this question, it is important to take nothing for granted. Just because something has always been taught does not mean that it's useful.


What are the challenges with this design?

It's important to identify early on what's going to be hard about implementing a given solution, and what can be done. For example, teaching investing skills is hard to do if no real money is involved. Teaching speaking skills is hard to do if the only source of feedback is a computer program.

Bad answers to this question tend to overlook issues of realism, motivation, and technical feasibility of interactions and feedback.

In response to answers to this question, always plan to ask "what else?"


What specifically will students do?

The answer to this question is the actual design. It specifies both what students do and what kind of feedback they get. Initial designs can just say "they'll drive a simulated car, encounter difficult weather conditions, and see what happens when they make different driving choices." Later designs need to give specific scenarios and what will be learned in each one, e.g., "they'll drive on an icy highway with traffic going too fast and encounter a several car pile-up; they'll learn how easily skids can occur and how to recover from them."

Bad answers usually just list topics, e.g., "bad weather driving," and fail to say what the student actually does, what choices there are, and what feedback occurs.


What support will students have to help them learn from their experiences?

This question looks at the support for learning that will be provided beyond the direct support for conducting a task. Some ways to support learning are to provide

Bad answers to this question can go to far in either direction. They can assume that no additional support for learning is necessary, that students will learn everything they need to know by direct experience (so-called "discovery learning"). Or they can provide too much support, i.e. by trying to tell or show a learner what they need to know, and shortchange them on opportunities to acquire and learn to apply new knowledge or skills themselves.