查看文章

Randomized controlled trials (RCT) have convincingly documented the efficacy of psychotherapy in the treatment of clients with various diagnoses and problems of life (Roth & Fonagy, 2005). RCT designs have thus played a significant role in establishing the efficacy of psychotherapy. However, treatment package/methods in RCT designs explain surprisingly little comparative outcome variance (Wampold, 2001; Wampold & Brown, 2005), while different aspects of psychotherapy process are consistently and convincingly associated with therapy outcome. These associations are demonstrated in both carefully controlled studies and in naturalistic studies (Orlinsky, Rønnestad, & Willutzki, 2004). We do not know the relative contribution of therapist effects and methods effects to the outcome of psychotherapy, as some studies have shown no therapist effects in RCT trials (eg Elkin, Falconnier, Martinovich, & Mahoney, 2006).), while others have shown rather substantial therapist effects (Blatt, Zuroff, Quinlan & Pilkonis, Lutz, 2006; Kim, Wampold & Bolt, 2006). Substantial controversies regarding the most appropriate statistical procedures to use need to be resolved in order to come to agreement on this important issue. However, there is agreement that therapist effects are best investigated in naturalistic studies (Crits-Christoph & Gallopp, 2006; Elkin, 2006, Wampold, 2006). Furthermore, successful therapies of various orientations are described similarly by clients, suggesting that there are some common elements in successful therapies (Lambert & Ogles, 2004). Analyses of studies published during the last 50 years, including almost 300 empirical …