illustrates the difficulties in transferring knowledge appropriately from one context to another (see Chapter 3). Despite these shortcomings, it is clear that this student is starting on the path to scientific thinking, leaving behind the more superficial “I’d take all the bad stuff out of the water” type of explanation. It is also clear that by making the student’s thinking visible, the teacher is in an excellent position to refine her (and perhaps the class’s) understanding.
Striking changes appeared in students’ scientific reasoning. In September, there were three ways in which the students showed little familiarity with scientific forms of reasoning. First, the students did not understand the function of hypotheses or experiments in scientific inquiry. When asked for their ideas about what could be making the children sick, the students tended, with few exceptions, to respond with short, unelaborated, often untestable “hypotheses” that simply restated the phenomena described in the problem: “That’s a thing…. Ah, I could say a person, some person that gave them something…. Anything, like give poison to make his stomach hurt” (Rosebery et al., 1992:81).
Second, the students conceptualized evidence as information they already knew, either through personal experience or second-hand sources, rather than data produced through experimentation or observation. When asked to generate an experiment to justify an hypothesis—“How would you find out?” —they typically offered declarations: “Because the garbage is a poison for them…. The garbage made the fish die” (Rosebery et al., 1992:78).
Third, the students interpreted an elicitation for an experiment—“How would you be sure?” —as a text comprehension question for which there was a “right” answer. They frequently responded with an explanation or assertion of knowledge and consistently marked their responses as explanatory (“because”): “Because fish don’t eat garbage. They eat plants under the water” (page 78).
In the June interviews, the students showed that they had become familiar with the function of hypotheses and experiments and with reasoning within larger explanatory frameworks. Elinor had developed a model of an integrated water system in which an action or event in one part of the system had consequences for other parts (Rosebery et al., 1992:87):
You can’t leave [the bad stuff] on the ground. If you leave it on the ground, the water that, the earth has water underground, it will still spoil the water underground. Or when it rains it will just take it and, when it rains, the water runs, it will take it and leave it in the river, in where the water goes in. Those things, poison things, you aren’t supposed to leave it on the ground.