Numerical landmarks are useful--except when they're not.

Placing landmarks on number lines, such as marking each tenth on a 0-1 line with a hatch mark and the corresponding decimal, has been recommended as a useful tool for improving children's number sense. Four experiments indicated that some landmarks do have beneficial effects, others have harmful effects, and yet others have no effects on representations of common fractions (N/M). The effects of the landmarks were seen not only on the number line task where they appeared but also on a subsequent magnitude comparison task and on correlations with mathematics achievement tests. Landmarks appeared to exert their effects through the encodings and strategies that they promoted. Theoretical and educational implications are discussed.