Testing the validity of a continuous false belief task in 3- to 7-year-old children.

In two studies, we examined young children's performance on the paper-and-pencil version of the Sandbox task, a continuous measure of false belief, and its relations with other false belief and inhibition tasks. In Study 1, 96 children aged 3 to 7years completed three false belief tasks (Sandbox, Unexpected Contents, and Appearance/Reality) and two inhibition tasks (Head-Shoulders-Knees-Toes and Grass/Snow). Results revealed that false belief bias-a measure of egocentrism-on the Sandbox task correlated with age but not with the Unexpected Contents or Appearance/Reality task or with measures of inhibition after controlling for age. In Study 2, 90 3- to 7-year-olds completed five false belief tasks (Sandbox, Unexpected Contents, Appearance/Reality, Change of Location, and a second-order false belief task), two inhibition tasks (Simon Says and Grass/Snow), and a receptive vocabulary task (Peabody Picture Vocabulary Test). Results showed that false belief bias on the Sandbox task correlated negatively with age and with the Change of Location task but not with the other false belief or inhibition tasks after controlling for age and receptive vocabulary. The Sandbox task shows promise as an age-sensitive measure of false belief performance during early childhood and shows convergent and discriminant validity.