The development of evidence evaluation skills

To assess developmental differences in evidence evaluation skills, 77 second- and third-grade students, 85 sixth- and seventh-grade students, 36 non-college-educated adults, and 40 college students were presented with four data sets depicting plants grown by each of four people. The data sets presented a perfect positive or zero correlation between plant health and the presence or absence of one variable, believed by participants to have a causal influence on growing healthy plants, or another, believed to have no causal influence. In each of three missing data conditions, the data sets depicted instances in which the status of the variable, outcome, or both were unknown in addition to the contingency data. After each data set was presented, participants judged (and justified) the causal status of the variable. Although demonstrating a basic competence, the two groups of children were more strongly influenced by prior beliefs and missing data than were the two adult groups. There were also age or educational differences in participants' tendency to justify judgments on the basis of the contingency data. The implications of the results for conceptualizing the continuity or discontinuity of children's, adults', and scientists' evidence evaluation skills are discussed.