often have the memory skills to either record information, record sufficient information, or consult such information when it has been recorded.

The importance of experience is highlighted by the results of studies conducted over several weeks with fifth and sixth graders. After several weeks with a task, children started making more exclusion inferences (that factors are not causal) and indeterminacy inferences (that one cannot make a conclusive judgment about a confounded comparison) and did not focus solely on causal inferences (e.g., Keselman, 2003; Schauble, 1996). They also began to distinguish between an informative and an uninformative experiment by attending to or controlling other factors leading to an improved ability to make valid inferences. Through repeated exposure, invalid inferences, such as invalid inclusions, dropped in frequency. The tendency to begin to make inferences of indeterminacy suggests that students developed more awareness of the adequacy or inadequacy of their experimentation strategies for generating sufficient and interpretable evidence.

Children and adults also differ in generating sufficient evidence to support inferences. In contexts in which it is possible, children often terminate their search early, believing that they have determined a solution to the problem (e.g., Dunbar and Klahr, 1989). In studies over several weeks in which children must continue their investigation (e.g., Schauble et al., 1991), this is less likely because of the task requirements. Children are also more likely to refer to the most recently generated evidence. They may jump to a conclusion after a single experiment, whereas adults typically need to see the results of several experiments (e.g., Gleason and Schauble, 2000).

As was found with experimentation, children and adults display intraindividual variability in strategy usage with respect to inference types. Likewise, the existence of multiple inference strategies is not unique to childhood (Kuhn et al., 1995). In general, early in an investigation, individuals focus primarily on identifying factors that are causal and are less likely to consider definitely ruling out factors that are not causal. However, a mix of valid and invalid inference strategies co-occur during the course of exploring a causal system. As with experimentation, the addition of a valid inference strategy to an individual’s repertoire does not mean that they immediately give up the others. Early in investigations, there is a focus on causal hypotheses and inferences, whether they are warranted or not. Only with additional exposure do children start to make inferences of noncausality and indeterminacy. Knowledge change and experience—gaining a better understanding of the causal system via experimentation—was associated with the use of valid experimentation and inference strategies.


In the previous section we reviewed evidence on developmental differences in using scientific strategies. Across multiple studies, prior knowledge

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement