Abstract

The article focuses on the meaning of "clinical meaningfulness" as opposite to "statistical significance", starting from a discussion on the probabilistic approach to the statistics, widespread in the experimental model of research. Some authors argued that the use of probabilistic statistics could give little benefit in the applicative fields of psychological research, advocating an effect-size approach that may overcome problems as the reduced size and/or high heterogeneity of the sample and the low power of the research. The paper reviews measures of effect-size for group comparisons, as Cohen's d, and indices of clinical significance as the Reliable Change Index, suitable for n = 1 studies. An example of use of the Reliable Change Index to test the efficacy of a clinical treatment for a patient with an anxious-depressive syndrome is presented.