Which single-subject design has the fewest threats to internal validity?

• A. A
• B. AB
• C. BAB
• D. ABAB

Answer: (D)

In this kind of question, the strongest protection against internal validity threats comes from designs that demonstrate the effect repeatedly under different conditions. A sequence that alternates between baseline and intervention, then withdraws and reintroduces the intervention, provides multiple demonstrations of whether the behavior changes in concert with the manipulation. When the target behavior shifts during the introduction of the treatment, shifts back when the treatment is removed, and then shifts again when the treatment is reapplied, it makes it much more likely that the observed change is actually caused by the intervention rather than by other factors such as maturation, history, or measurement drift.

This repeated reversal serves as replication within the same subject, helping rule out alternative explanations. Other designs that only include a single baseline or a single introduction of treatment offer less opportunities to demonstrate that the treatment is the causal factor, so they are more vulnerable to threats to internal validity. In short, the design that uses multiple baseline–treatment reversals provides the clearest evidence that the intervention is producing the observed change, and thus has the fewest internal validity threats. Ethical note: withdrawing effective treatment can raise concerns, so consider the behavior and context before applying reversals.