Our best learner was a conservative with mostly moderate opinions, whereas one of our next best learners was a political-social liberal.
There were differences in reasoning along at least three dimensions that we examined. First, there were differences in complexity indicated by several different measures. Dave, George, and Mitch were high in complexity; Eileen, Robbie, and Jen were low in complexity. To some extent, students who showed more complex reasoning were those who had learned more, and thus had more information with which to display complex reasoning. However, this coupling of learning with complexity must be qualified. Eileen was a better learner than Dave on all measures but Dave's reasoning was more complex. Factors beyond learning contribute to complexity. A certain amount of learning, we suggest, is critical for displaying complex reasoning; thus we would not envision any additional complexity from someone like Robbie. Dave's higher complexity, a willingness to consider larger parts of the story in responding, is something one might not have expected based on his average learning. But average learning, in our case, is really quite a bit of learning. The students learned a lot about the story, and their ability to do more or less with that learning is partly a matter of some additional factor - an approach to our tasks, a disposition regarding questions without "right answers," or other individual difference factors.
Second, there were differences in stability and text influence on reasoning. All students allowed what they read to influence their responses to reasoning probes, including questions about fairness. But Eileen and Dave were influenced more than Jen; and Robbie, Mitch, and George were quite stable in their opinions over different texts. It is interesting that the degree of influence was not related to amount of learning in any obvious way; the one who learned most, George, and the one who learned least, Robbie, were both text-resistant. The students showing the most influence were among the best learners. We suggest the more general relationship here is that more learning leads to more text influence in reasoning, even in questions of value, provided the learning is specifically informative for the questions being asked. George is an understandable exception: He appeared to adopt a critical attitude toward the texts; he also appeared to take some pride in his own abilities and knowledge, adding to his text resistance. Our hunch is that it is more typical, under conditions of low knowledge, to allow information presented in texts to affect reasoning on questions for which it is relevant.
Third, part of the difference we observed about students' reasoning on fairness issues - fairness to Panama and Colombia; did the United States do the right thing - could be attributed to differences in the way students thought about fairness in these contexts. Students from a relatively homogenous background, as these were, can be expected to have a common ethical value to apply to everyday issues of fairness. In international relations, however, the idea of fairness can be interpreted in legalistic, financial, national sovereignty, and other perspectives. Three students - George, Mitch, Robbie - took a strictly legalistic approach to fairness; Jen and Dave took a noninterventionist approach, one that placed value on a nation's right to be free of outside interference. (Their differences have obvious parallels in discussions of basic foreign