“…First, symptom-or problemfocused goals may simply be harder to attain than goals from other categories, as pointed out by Berking et al (2005), which should be reflected in lower effect sizes. Second, patients with particular monosymptomatic disorders, such as phobias, panic disorders or somatoform pain disorders, have repeatedly been shown to prefer symptom-related therapy goals (Dirmaier et al, 2006;Grosse Holtforth et al, 2009;Schöttke et al, 2011;Trachsel et al, 2008), which might not be adequately operationalized with a multidimensional measure of treatment outcome like the FEP. In this case, the effect sizes would turn out to be numerically small due to the measurement of irrelevant constructs.…”