Abstract
This dissertation explores capabilities o f high school science students for reasoning about representational displays of data. Like Galileo, the students were attempting to understand astronomical observations. On the one hand, the students had the advantage of using the latest tools and data of astrophysicists. These included high-resolution digital images created by attaching very sensitive digital cameras to telescopes, revealing objects far fainter and in greater detail than could be seen through Galileo’s telescope. The students also used image enhancing computer software as a tool for displaying and analyzing the digital image data. On the other hand, these powerful tools have vastly compounded the potential for being deceived by appearances. Indeed, this is a problem not just for students and scientists, but all of us, whenever we try to make sense o f computer-generated displays of data. Recently, such displays have accompanied front-page newspaper articles, as shownin Figure1-1. So it is timely to explore ways for students to develop their capabilities for interpreting informative, but deceptive computer-generated representational displays o f data. The students’ substantial capabilities for interpreting image displays made this study possible. When high school physics and astronomy teacher Richard Lohman and I first introduced image-processing activities, we were struck by the variety of interpretations that were invented by his students. Over the years, we sought ways to help students build upon their capabilities for interpreting image displays. We tried to create image-processing based activities where students’ inventiveness was valued over their ability to learn any particular methods. We tried to find image data that could be interpreted in different ways. We often encouraged each student to create multiple methods. And we tried to ensure that students had adequate resources for creating and critically evaluating their methods. In these ways I learned how to engage students in articulating and evaluating interpretations o f a variety o f image data in the context o f various data analytic tasks. Analyzing a corpus of student-generated interpretations, I’ve identified intellectual resources that contribute to their interpretive capabilities.
Not surprisingly, students’ perceptual capabilities may be their most fundamental resource for interpreting the visually rich image displays. In particular, students frequently focus their attention on particular visual attributes, such as those identified in Figure 1-2, and base their interpretations upon what the visual attributes suggest. A few examples will illustrate the range of students’ capabilities for interpreting visual attributes. I begin with examples demonstrating the best of the students’ work.
