Prevalent practice in the literature of using correlated common and specific factors as independent predictors in classical ANOVA models is both statistically unsound and conceptually distorted

The misleading Dodo Bird verdict. How much of the outcome variance is explained by common and specific factors? Giulio de Felice et al. New Ideas in Psychology, Volume 54, August 2019, Pages 50-55. https://doi.org/10.1016/j.newideapsych.2019.01.006

Abstract: The literature on psychotherapy research makes use of the so-called "Dodo Bird Verdict" to show that therapeutic change owes more to common factors than to specific techniques. According to the bulk of the empirical literature, common factors explain 30–70% of therapy outcome variance, while specific factors account for between 5% and 15%. This formulation is based on the assumption that common and specific factors are independent of each other. The present study uses a systematic review of the literature to empirically demonstrate that common and specific factors of change are actually correlated. In other words, the prevalent practice in the literature of using correlated common and specific factors as independent predictors in classical ANOVA models is both statistically unsound and conceptually distorted. We offer several alternative proposals for a sensible re-evaluation of the Dodo Bird verdict.