Despite evidence that young children are sensitive to differences in angle measure, older students frequently struggle to grasp this important mathematical concept. When making judgments about the size of angles, children often rely on erroneous dimensions such as the length of the angles' sides. The present study tested the possibility that this misconception stems from the whole-object word-learning bias by providing a subset of children with a separate label to refer to the whole angle figure. Thirty preschoolers (M = 4.86 years, SD = .53) were tested with a pretest-training-posttest design. At pretest, children showed evidence of the whole-object misconception. After training, children who were given a novel-word label for the whole object improved significantly more than those trained on the meaning of "angle" alone. (author abstract)