A meta-analysis of the contribution of eye movements in processing emotional memories

In therapy1

In 15 clinical trials, EMDR with eye movements beat EMDR without them by a moderate, significant margin (d = 0.41).

In the lab1

In 11 laboratory studies of recalling a distressing memory with and without eye movements, the effect was large (d = 0.74).

Vividness most of all1

The biggest difference was in how vivid the memory felt, in the non-therapy studies (d = 0.91).

Fidelity mattered1

In the therapy studies, how faithfully EMDR was delivered moderated the effect of eye movements.

Why redo it1

A previous meta-analysis found no significant eye-movement effect; the authors argued its methods may have caused a type II error, missing a real effect.

Two kinds of evidence1

Clinical dismantling trials and laboratory studies were analysed separately rather than pooled.

What it implies1

The authors read the results as support for theories that EMDR works by processes different from other exposure-based therapies.

Working memory2

One leading explanation: a dual task while recalling takes up working memory, making the memory less vivid and intense.

An outlier2

Most meta-analyses find adding eye movements makes little or no difference; this one used fewer exclusion criteria.

Allegiance2

Meta-analyses describe a high risk of allegiance bias in EMDR studies.

The Role of EMDR Therapy in Medicine

Shapiro's 2014 review, which cites this result alongside 20 trials of the eye movements.

Effect size

What d = 0.41 and d = 0.74 mean, and how to read them.

EMDR

Where this result sits in the wider evidence.