Measuring the intervention in effectiveness research.

Inconsistent implementation of the intervention in the field setting presents a threat to the validity of the conclusions of an effectiveness study. Inconsistent implementation results in variability in the actual dosage of the intervention received by the participants, which leads to variability in outcomes achievement and, consequently, to Type II error. This article discusses the methodological implications of inconsistent intervention implementation and advances a strategy as a solution to this problem. The strategy proposes to quantify the dosage of the intervention as a continuous variable and to use this variable in the statistical analysis. The benefits of this strategy are illustrated with an empirical example from the Self-Help Intervention Project.