Sometimes a design with alternating treatments may be attractive for psychological reasons. For example, in their Experiment I with eight subjects, Ingham and Andrews (1973) studied two methods of therapy for stutterers. In Treatment A (token reward method), subjects were given tokens that had value like money for having a low or reduced percentage of stuttered syllables in a 45-minute therapy session. In Treatment B (token reward and penalty method), they were given tokens for appropriate speech behavior but lost tokens after an increased rate of stuttering. Ingham and Andrews alternated blocks of sessions with these two treatments in order to learn whether one reward structure led to superior performance compared to the other. Inspection of a graph of average performance in this experiment suggests that Treatment B was more effective than Treatment A in reducing stuttering.