 Well, you've made it this far so we're going to add on another layer. It's an A, B, A, B design. Again, it's single subject, it's experimental, and this time we have four phases. A, baseline, B, intervention, back to baseline, yes, and back to an intervention phase. We have the same criteria as previously. All conditions require stability before you change. We can compare the A's to the A's and the B's to the B's. An A, B, A, B design is often referred to as a reversal or a full reversal design. Under certain conditions it may also be referred to as a withdrawal design. The determining factor of whether or not it's a reversal or a withdrawal has to do with the type of intervention that you're using, the type of stimuli, or the type of conditions for the B. We could use classroom management procedures. You can start with your traditional procedures and then we could implement a token economy. Then we could remove that token economy and then go back to it, allowing us to compare the token economy conditions to the baseline conditions to get an idea of which was more effective.