Dodo bird verdict

Few differences in head-to-head trials1

Since Luborsky's 1975 review, many comparisons of bona fide therapies find small or no differences.

Small treatment differences2

Where differences exist they are at most about Cohen's d = 0.20, far smaller than the effects of alliance or empathy.

Techniques as vehicles1

Wampold and colleagues (2009) argued that techniques matter mainly as ways of activating common factors.

The original race1

In Alice in Wonderland the wet characters run round in a race to get dry; nobody measures how far or how long each ran, and the Dodo rules: "Everybody has won and all must have prizes."

What the shared factors are1

Supporters mean things like a relationship with a therapist who is warm, respectful and has high expectations for the client's success.

One study tipping the balance1

Wampold found that a CBT advantage over rational therapy for panic disorder rested on one study with an unusually large effect (1.02); without it, the difference was not significant.

Comparing like with like1

Crits-Christoph found that 79 of the 114 articles in a 1997 meta-analysis supporting the verdict involved similar comparisons, which can hide real differences.

Uneven contests1

Some trials compare a therapy against a waiting list or medication, or give one side's counsellors only a few days of workshops.

Treatments that can harm1

Negative responses have been reported in some substance-abuse work, some types of grief therapy and certain techniques with trauma patients.

What hangs on it1

If the verdict holds, clinicians could feel free to use any therapy they see fit; if not, they would likely have to use empirically supported ones.

Anxiety disorders1

The strongest counter-evidence: cognitive behavioral therapy shows larger effects than alternatives for several anxiety disorders.

Some treatments harm1

Lilienfeld (2007) and others note some therapies can harm, which "all are equal" cannot be true of.

Flawed meta-analyses1

Hunsley and Chambless argue the verdict rests on biased study selection and overgeneralizes from limited comparisons.

Policy and training1

The answer affects funding and whether clinicians must use specific empirically supported treatments; see Evidence-based practice in psychology.