Is tracing or copying better when learning to reproduce a pattern?

Learning to write requires the repeated manual production of spatial patterns. It remains unclear whether tracing or copying provides better training: tracing provides accurate and immediate performance feedback, whereas copying may require greater use of memory and recall during training. We asked sixteen adults to copy or trace novel patterns then reproduce these from memory using a stylus and tablet PC. A week later, a retention test was performed. Sophisticated analyses indexed the extent to which participants had learned the dimensions and shape of patterns. We found that participants: (a) showed better shape and dimensional accuracy when tracing; (b) had better shape and dimensional retention immediately after tracing; (c) showed no differences between copying and tracing in their ability to redraw the pattern (shape or dimensions) 1 week later. Our methods provide a useful starting point for examining training and feedback on the generation and recall of spatial patterns.