Abstract
In this thesis, I argue for the important role of human perception in defining the concept of image, central to geometrical optics. Further, I advocate that students be taught to use a specific type of model of a human observer, which I call a "transparent detector model." This is a runnable model with a simple causal mechanism for interpreting rays of light, and serves to procedurally define the concept of image. More generally, I propose that transparent detector models may facilitate the learning of concepts in many domains of physical science. In an instructional study, I examined the instructional effect of one such model of the human visual system. This model, embodied in a "Plastic Person" artifact, uses the divergence of light rays as input, and infers the location and distance of the source of these rays. The inferencing mechanism of the detector uses the kinesthetic feedback on the positions of the eyeballs in their sockets to determine an object's location. Students aged 14-15 were taught the fundamentals of image formation either using the detector model, or using a more traditional approach in which image is defined geometrically as the crossing point of multiple rays. The latter version of instruction included an observer, but one without an explicit inferencing mechanism. The results show that students who were taught a runnable model of visual perception exhibited a better understanding of the notoriously difficult relationship between an observer and a virtual image. They were also better able to identify the location of an image in a range of real-world optical situations, and were less likely to think of it as located on the surface of a mirror or lens. Students in both instructional groups showed significant learning in five out of six areas of well-known student difficulty. This suggests that including an observer in either form within their instruction was highly effective in increasing students' understanding.
