Predicting that when presented with a general context children and adults would produce different specific inferences, a study examined children's and adults' script representations of a common event and the use of representation in comprehension. In the first phase of the study, 23 fourth grade and 48 college students were asked to generate actions that typically occur while buying lunch at the school cafeteria. In the second phase of the study, 30 fifth grade and 30 college students listened to a short passage based on the generated scripts and were later given a recognition task. Each subject was handed a booklet containing 16 sentences, one sentence per page. Each sentence was one of four types: verbatim, unrelated, or one of two types of related foils--changed or new. There were 8 sentences stated verbatim in the story, 3 sentences unrelated to the lunchroom story, and 5 related foils--2 changed foils and 3 new foils. The changed foils were sentences presented in the story but changed slightly so that the sentence was made to be either congruent with the children's lunchroom script or congruent with the adults' lunchroom script. New related foils were sentences not explicitly stated in the lunchroom story but containing actions consistent with either the adults' or children's lunchroom script. Unrelated foils were sentences having no relation to the story. Findings revealed that children rated foils that were congruent with their representation as more likely to have been stated in the passage than foils that were congruent with the adults' representations. The adults showed the opposite tendency. (HOD)