Proceedings of the Annual Meeting of the Cognitive Science Society

East Asian students routinely outperform American peers in mathematics. One source of this learning gap may be linguis-tic, such as explicitly naming part-whole relations in fractions (e.g., of four parts, one in Korean vs. one-fourth in English).Our study examined whether adopting such language would improve American children’s number-line estimates. To testthis, 83 10-year-olds were read fractions using either Korean-style or English names over pretest, training, and posttest. Inboth conditions, number-line problems either had no landmarks, landmarks that matched the denominator, or landmarksthat did not match the denominator. As expected, we observed a session by problem type interaction (F=2.71, p¡.05),indicating that feedback improved accuracy most for problems involving matching landmarks. Surprisingly, the effect ofKorean naming was to reduce accuracy across all problems and test phases (ps ¡ .01). Results offer an important warningagainst linguistic reform that may be harmful for American students.