15
A-B-A design
a type of single-case design having three phases: a baseline condition in which no treatment is present (Phase A), a treatment condition in which manipulation is introduced (Phase B), and a return to the no-treatment condition (Phase A). The design allows for evaluation of the introduction of the treatment by comparing the dependent variable between the first two phases (A-B sequence) as well as evaluation of the removal of the treatment by comparing the dependent variable between the last two phases (B-A sequence). This greatly reduces the possibility of a coincidental treatment effect, which may occur in the simpler A-B design.