Pick an independent variable that you are interested in. Then explain how this variable could be manipulated using a between subjects design, and then using a within subjects design. It's fine if your example is strange (not what you'd do in real life); the point is to show that you understand between vs. within subjects designs