Abstract
This research focused on young children's abilities to use landmarks when they lost their senses of where the target is located. In this experiment, a target object was hidden under one of four identical boxes arranged in a 2 × 2 matrix. Two dolls served as landmarks, one (the direct landmark) was behind the target box and the other (the indirect landmark) was behind the box in the diagonal position. The entire table top was rotated behind a screen, so that children had to infer the position of the hidden object from the locations of the landmarks. In addition, as children couldn't see the landmark behind the boxes, they had to infer the position of the hidden object from the other visible landmark. The children performed such an inference task significantly better than chance. Even if the direct landmark was occluded from their view, they could infer the target position from the indirect landmark. However, in some condition they failed to use the indirect landmark, which was interpreted as revealing their failures to become aware of the change of the situation. Taken together, this research indicated that 3-year-old children had the ability to code the direction of the target relative to another object. It was interpreted that their difficulties of using the indirect landmark were derived not from the coding problem per se, but from their inability to start retrieving relevant pieces of information.
