Re:Request for Ricky

I thank Ricky and Louise for calling my attention to some details of the Davidson & Parker meta-analysis that I had forgotten.

Davidson & Parker do indeed state that "the EMDR-EFixDR [eyes fixed desnsitization and reprocessing] effect size is marginally significant if one examines only clinical populatins satisfying DSM diagnoistic criteria" (p. 311).

I will return the favor to Ricky and Louise by reminding them of the discussion that follows this statement. It is important to note that effect is not statistically significant (i.e., p < .05) but "marginally" significant (p > .05). This speaks to the issue of low power that Sandra has previously raised. Power is a function of the the magnitude of the effect size and sample size: the smaller the effect size, the larger the sample required to detect a difference. By convention, we think of a study as being adequately powered when a study has power of .80 or greater.

Based on the effect sizes obtained in the meta-analysis, looked at the issue of power in two ways. First, they computed the necessary sample size to necessary conduct a single study comparing EMDR with EFixDR with a power index of .80, given the liberal assumption of alpha = .05, ONE-TAILED (rather than two-tailed). The study would require a sample of 144 subjects.

Second, they determined the number of new studies that would have to be conducted in order to find within the meta-analysis a significant difference. The result was 38 studies.

So, both of these indexes indicate that, if there is an effect there, it is relatively small compared to the remaining EMDR package. Moreover, it should be pointed out that even if someone actually conducted the large scale study necessary or if 38 more studies were conducted, there is no guarentee that the results would favor EMDR over EfixDR. It just means that, IF there is an effect to be found, THEN the studies would have adequate power to detect that effect with a reasonable degree of certainty. The small magnitude of effects in the published outcome data certainly offer a stark contrast with the large effects that people often talk about based on their clinical experience.

I see very little reason here to change my overall conlcusion about the effects of eye movements on outcome for EMDR: that is no convincing evidence showing eye movements contribute to outcome of EMDR. If I were to ammend it, it would be something like "there is no convincing evidence showing a large effect of eye movements on outcome of EMDR. Perhaps there is a small effect of eye movements, but this has yet to be convincingly established. More research is needed."

Replies:

There are no replies to this message.

| Behavior OnLine Home Page | Disclaimer |

Copyright © 1996-2004 Behavior OnLine, Inc. All rights reserved.