Inhibition of the knowledge of natural numbers during the processing of fractions: A priming study in adults.

Learning and processing fractions is complex for children and even adults. Errors produced by children suggest that the knowledge of natural numbers can interfere with this learning and processing (e.g., Stafylidou & Vosniadou, 2004). Here, a priming paradigm was used to test this interference and its control in adults. If adults rely on the knowledge of natural numbers to process fractions, performance should be poorer to compare fractions with common numerators (incongruent pairs because the larger fraction is made up of the smaller denominator) than to compare fractions with common denominators (congruent pairs because the larger fractions is made up of the larger numerator). If adults inhibit the knowledge of natural numbers to correctly compare fractions with common numerators, comparing natural numbers should be more costly after the processing of these fractions. Adult participants were presented pairs of fractions with common denominators or numerators, pairs of natural numbers and pairs of squares and were asked to perform a magnitude-comparison task. The comparison of fractions, natural numbers and squares primed the comparison of natural numbers. In the general priming condition, the fraction components differed from the natural numbers that were presented as the probe. In the specific priming condition, they were the same. A questionnaire assessed the conceptual and procedural knowledge of fractions. The results showed that response times were slower and the percentage of errors was higher for fractions with common numerators than for fractions with common denominators, which suggest that the knowledge of natural numbers interfered with the processing of fractions with common numerators. Furthermore, participants compared more slowly and less accurately natural numbers when they were primed by fractions with common numerators compared to when they were primed by fractions with common denominators. This negative priming effect was larger in the specific priming condition than in the general priming condition. These results suggest that the knowledge of natural numbers was inhibited during the processing of fractions with common numerators. In conclusion, the knowledge of natural numbers interferes with the processing of the magnitude of denominators in adults and this interference has to be actively controlled in order to correctly compare fractions