This study investigates whether children can process analogies with better performance for visualized than for verbalized formats. Fifth and sixth graders were presented texts explaining procedures for drawing polygons by dividing up a disk into either five, six or eight sectors and joining the end points of those sectors by line segments. Then they were to analogize and point out task-appropriate figural aids for drawing a right triangle and a square each inscribed in a separate disk: The participant was to understand how a square was drawn by dividing up the disk into four sectors and joining their end points. Predictive importance of those geometry skills and analogy test remarks were assessed for identification of figural aids to draw triangles and squares. Results showed that fifth-graders' analogy test remarks predicted their choice of figural aids in drawing triangles and squares. For the sixth graders, facilitative effects of inserted question on their identification of figural aids were obtained, while just the reverse was the case for the fifth graders. The results were interpreted in terms of thought levels of Van Hiele.