Researcher allegiance as hidden moderator in psychotherapy outcome research

Randomized controlled trials that compare 2 rival treatments are used as the standard design to identify those constituents of complex psychotherapeutic treatments that critically impact symptom improvement. This paper examines the impact of unbalanced researcher allegiance, i.e. the researchers' preference regarding a particular treatment, as an extra -therapeutic moderator of the treatment effect. As researcher allegiance can only be measured at study level, meta -analytic strategies are required to investigate the association between researcher allegiance and outcome. Our review shows that researcher allegiance has a considerable impact on outcome. Despite the finding that researcher allegiance may impact outcome particularly in studies of poorer quality, other causal pathways need to be considered when trying to understand the relationship between unbalanced researcher allegiance and outcome. Although the association between unbalanced researcher allegiance and outcome is well established, the causal pathways that mediate this association are not well understood to date. Accordingly, we conclude that explicit statements by researchers regarding their preferences in a comparative outcome trial (e.g. by stating clear hypotheses or by including a conflict of interest statement) appear to be the most straightforward response to our findings in order to protect the validity of conclusions from psychotherapy outcome research.