Previous research has verified the benefits obtained when learners trace out worked examples with the index finger. Our study conducted two experiments to explore the reasons for this phenomenon and its generalizability. Experiment 1 compared the learning effects among tracing, non-tracing, and cueing methods. The cueing method was included to isolate the variable of drawing attention. Students employing the tracing method obtained higher transfer test scores compared to those employing the cueing and non-tracing methods, rating the far transfer test as easier. The tracing effect was verified to have a greater impact on learning than solely focusing learners' attention. Experiment 2 compared the learning effects among three methods -- tracing with the index finger, tracing with a computer mouse, and observing others tracing. Students who merely observed tracing obtained lower far transfer scores than those who traced either with their finger or with the mouse, rating the tests as more difficult. There was no significant difference between the other two groups. Tracing promotes learning more than merely observing tracing. The learning benefits of tracing partly stem from the fact that it is action-based and can be generalized to mouse tracing.