◆Design experiment approach◆

 Characteristics and research cycles of the design experiment approach

 Oshima (2004) points out characteristics of the "Design Experiment Approach" as shown in the five points below in his paper that explains the approach in the same issue of Journal of the Japanese Society for Information and Systems in Education, in which Miwa and Saito (2004) appeared.

1. To design learning in a chaotic situation rather than in a scene elaborated in a laboratory


2. Not to focus on a single dependent variable, but to handle more than one variable at the same time


3. Not to verify a hypothesis, but to create a profile for a designed environment


4. Not to follow the established procedure, but to modify it according to circumstances


5. To deal with learning in a social situation, rich in intellectual resources

 Oshima proposes we accept the chaotic situation of a "classroom" as it is, rather than simplify it as much as possible so that laboratory researchers accept it for their analyses. He argues researchers should observe what will happen in a multifaceted way without sticking to rigidity so that a profile which is directly useful for the next action can be obtained. His idea is so bold that traditional learning psychologists could criticize it as irresponsible.

 Figure 2 shows a research cycle of the Design Experiment Approach proposed by Miyake and Shirouzu (2003). Based on a learning model which indicates the level of learning and the expected level of achievements, a lesson is to be designed and implemented, and then the outcome is observed. Although researchers cannot get conclusive evidence to show a causal relationship, the "learning model" presented at first might be close to the truth, if similar phenomena occur several times. They argue the most "ordinary approach" is to improve the first model, work out a stronger design based on the model, and find out a principle for the most versatile design.

That's what I wanted to hear. It is exactly what instructional design illustrates, don't you think so?

Figure 2: Research approach of learning sciences (Miyake & Shirouzu, 2003, p.71)