Building Functional Models: Designing an Elbow.

Penner, David; Giles, Nancy; Lehrer, Richard; and Schauble, Leona.

Journal of Research in Science Teaching. Vol. 34. No. 2. PP 125-143. (1997)

Abstract prepared by Chuck Downing, PhD.

The idea of modeling or model building is a topic at least touched upon in nearly all science classrooms. The effectiveness of model building as a learning device depends on how effectively student embrace the true meaning of "modeling."

Most students believe that "science" consists of observing and recording-scientists make observations and accurately record those observations. According to student perceptions, keeping complete records is the key to scientific discovery. Students also believe that observing is "non-participatory"-models to them are copies of physical phenomenon.

However, most working scientists are model builders-their lives tend to be dominated by building and testing models. Scientists know that models channel observations and drive interpretations.

For the purpose of this paper, models can be defined in two ways:

1. The model looks like the "real thing."
2. The model performs the correct function of the "real thing."

Penner, Giles, Lehrer, and Schauble chose the problem of how the human elbow works to be investigated via model building. Before beginning model building, students did guided research. The students determined that not all mammalian "elbows" are the same in either form or function. This was a key finding, since it invited a more open-ended approach to student model building.

Key points to emphasize when teaching through model building are:

• Models are not merely representations of something.
• Evaluation and change are important in model building, and not just for the fixing of mistakes.
• Model-based reasoning takes time to develop.

For this study, a variety of materials were made available to students for use in model building. Cardboard tubes, Popsicle sticks, modeling clay, Styrofoam balls, hinges, springs, wooden dowels, flexible strips, tongue depressors, balloons, rubber strips, door hinges, posterboard, and glue were among the materials made available. Some items were deliberately included because of their match with structural features (e.g., dowels could represent arm bones). Others were included because they matched with perceptual features (e.g., Styrofoam balls matched the "bump" at the base of the elbow).

Students were given three 1-hour sessions on consecutive days to complete "more than one solution" to their model building problem. After building initial models in groups, results of these group projects were presented to the class. Time was given to revise or modify designs as groups desired.

Two groups of students were interviewed after the model building sessions: 1) model builders and 2) non model builders. During the individual student interview, each student was shown four "sample" models: two tongue depressors in a clay ball, a flexible drinking straw, a drawing of the elbow, and two pieces of posterboard hinged together with a string running through them. Ratings were given to each model in answer to a question on the functionality of the model (1 being "did not like" to 5 being "liked a lot").

There was some similarity between groups in rating the sample models. For example both groups liked the cardboard/string model best. There were differences between groups as well-model builders were much less pleased with the picture than were non model builders of similar age.

The most significant finding regarding the model-rating was in the types of critique-comments provided by students. There were statistically significant differences between the two groups-model builders were more likely to provide justification with a constraint than non model builders. In other words, model builders were better able to explain what the limitation of a model was than were non model builders.

As an example of this type response, consider the flexible drinking straw model. Based on their model building experiences and subsequent revision of models based on class discussion, model builders disliked this model because it did not limit backward-bending motion "like a real elbow." In addition, non modelers of the same age as the model builders universally failed to see why lack of functionality (in this case backward bending of the joint) was a major flaw in a model.

The study focuses on children's attempts to understand the function of their elbows through a process of model-based design. Via discussion, model building evaluation, and revision, children came to understand that not only motion, but also constraints on motion were important qualities to include in their models. […] In comparison to a nonmodeling peer group, modelers were largely able to ignore perceptual qualities when asked to judge the functional qualities of models. Further, they showed an understanding of the modeling process in general that was similar to that of children 3-4 years older. - p. 141

Oh, did I mention that the students in the study were first and second graders?

I suspect you are wondering how a study on early elementary students has any application to middle and high school students. Here's how: "In fact, the ideas held by middle and some high school students [reported in previous studies] seem to be the same as those held by our nonmodeling second graders [emphasis mine]. They shared a propensity to evaluate models by assessing their perceptual similarity to the phenomenon of interests." - p. 141

So what's the bottom line here?

I would hope that we want our students to function beyond the second grade level in as many areas as possible by the time they finish our classes. If model building and testing are central constructs to the process of science, we need to expose our students to those methods. However, simply saying "build a model" is not enough. There needs to be a sequence of "discussion, model building, evaluation, and revision" where students are allowed to explore multiple models that reflect both structural and functional qualities of whatever is being modeled.

I encourage you to consider where in your curriculum you provide time for students to proceed through a pathway like the one described above. If you find such pathways missing, where could one be included?

When you include model building in your curriculum, don't let your students off the hook when it comes to critiquing models designed by themselves or classmates-be sure to ask about the constraints of the model presented. You can use as an example of such critiquing that, while a flexible drinking straw is a model of the human elbow, it is limited in effectiveness because it allows for flexion in any direction, and the human elbow allows flexion in only one direction.

Science Education Reform Index

Let's Collaborate Index