A total of 512 children in Grades 1 through 6 received a conditional inference task using causal conditionals (If cause P, then effect Q) and a generation of alternatives task. The inference task used premises for which there were few or many possible alternative causes. Results show a steady age-related increase in uncertainty responses to the two uncertain logical forms, affirmation of consequent (AC) and denial of antecedent (DA), and an increase in production of disabling conditions for modus ponens. More uncertainty responses were produced to AC and DA with premises with many possible alternatives. Individual differences in inference production were related to numbers of alternatives produced in the generation task. Results support the idea that both developmental and individual differences in reasoning can be at least partially explained by differential access to knowledge stored in long-term memory.

The present article examines 2 predictions concerning conditional reasoning in children derived from Markovits's model of conditional reasoning. The first claims that children under 12 years of age should be able to respond correctly to uncertain logical forms if the premises and context enable them to access pertinent counterexamples from memory. The second concerns the effect of reasoning in a fantasy context. Previous studies have established that young children can correctly respond to certain reasoning problems with empirically false premises when these are presented in a fantasy context. However, this model of reasoning predicts that presenting empirically true premises in a fantasy context should decrease performance on the 2 uncertain logical forms. In Study 1, a total of 48 8‐year‐olds, 78 10‐year‐olds, and 74 12‐year‐olds were given 4 reasoning problems involving familiar premises. These problems were embedded in either a fantasy or a realistic context and presented via video tape. Results were consistent with the predictions made. Study 2 attempted to determine whether these results could be due to context or problem formulation. A total of 40 7‐year‐olds and 46 8‐year‐olds were given reasoning problems with either no context or with a visual image preceding the problems. Results showed that children did equally well in these conditions, and that providing an image did not improve performance.

No abstract