Monday, 15 October 2007

How experimenters influenced participants in the ganzfeld parapsychology experiment

An analysis of conversations that took place during ganzfeld parapsychology experiments has revealed researchers may have exerted an influence on their participants.

Ganzfeld experiments involve a 'sender' trying to project images from a video clip to a 'receiver' who is incubated, blindfolded, in a sound-proof room. The 'receiver' reports the images they believe they are receiving to a researcher who notes them down. Crucially, the next stage involves the researcher reviewing these images with the 'receiver', before the 'receiver' attempts to identify the video clip seen by the 'sender' from among three decoys.

Robin Wooffitt analysed recordings taken from ganzfeld experiments held at the famous Koestler Parapsychology Unit in Edinburgh during the mid 1990s. He found that as the researchers reviewed the images reported by the 'receivers', they tended to respond in two distinct ways.

After a clarification by the 'receiver', researchers sometimes said “okay” and moved decisively onto the next item. Other times, however, they said “mm hm” with an inquiring tone. After hearing this, 'receivers' typically tried to expand on their description, and as they did so, often ended up casting doubt on the clarity of their own imagery.

Wooffitt said that a researcher's choice to respond with “okay” or “mm hm” might seem inconsequential, but in fact the latter utterance clearly had an effect on the 'receivers'' confidence in their imagery. Consequently, he said, “it is at least possible that they [the 'receivers'] will have less confidence in relying on their imagery to identify significant events or themes in the video clips.”

If the researchers did influence participants in this way, could it help explain why sceptical researchers have tended to report more negative results than believers? Wooffitt told the Digest: "I think it has much more to do with the nature of one’s interactional style, and that doesn’t necessarily correlate with either sceptical or ‘pro-paranormal’ beliefs."

Wooffitt added that by furthering our "understanding of the impact of the social dynamics in psychology experiments more generally", his observations have implications beyond parapsychology research.

Wooffitt, R. (2007). Communication and laboratory performance in parapsychology experiments: Demand characteristics and the social organisation of interaction. British Journal of Social Psychology, 46, 477-498.

Post written by Christian Jarrett (@psych_writer) for the BPS Research Digest.


Annalisa Ventola said...

I'm a big fan of the BPS Research Digest, but I think that the title of this article is misleading. I haven't read the original paper for myself since I do not get the journal. However, it my understanding that Ganzfeld experiments are double blind. As I understand it, Wooffitt has done a psychological study of the interactions between experimenters and participants in this setting. There is nothing in this article (other than the title) suggesting that experimenters biased participants responses in any particular direction when it came to selecting target images.

Unknown said...

The finding is that experimenters' verbal responses affected the confidence of 'receivers' in their own imagery. This is a flaw in the experimental design. It is possible (though this was not tested by this study) that more sceptical researchers may have been more prone to making the kind of verbal responses that sowed doubt in the 'receivers'.

Unknown said...

Annalisa, I've made a change to the title, so that it is more explicit.

Anonymous said...

The experiments Wooffitt is referring to were double blind, meaning the researcher was blind to the target image. The researcher could hum and haa as much as he or she liked, it would make absolutely no difference. There is still only a 1 in 4 chance of choosing the correct image.
Come on guys, this is kindergarten stuff.

Unknown said...

Anonymous you are mistaken. Your analysis pre-supposes that there is no experimental 'telepathic' effect. If you 'know' there is no such effect, then there is no point conducting an experiment. If, however, the aim of the experiment is to test whether there is an effect or not, then these interactions are relevant. By sowing doubt in the receiver's confidence in his/her imagery they could increase the likelihood of a type 2 statistical error - that is, failing to observe an effect when in fact there is one.

Unknown said...

PS. Just to clarify - what I just explained still stands whether the experiments were double blind or not.

Beacon Schuler said...

Anonymous - the umming and ahhing is having an effect on the subjects responses. What the research shows is that noise is being introduced into the experiment by the way in which the experimenters interact with the subject. This is occuring whether the experiment is double blind (and the noise is unrelated to the desired response) or not.

Anonymous said...

There seems to be some confusion.
Skeptics on another website have pounced on this article claiming that it explains the positive results.
I was pointing out that it clearly does not.
A researcher sitting in with the receiver creates noise and can of course influence the outcome but only if anything to reduce the effect not increase it.
I'm sure the researchers were well aware of this.

Anonymous said...

Digest - this is not a flaw in the experimental design. As far as I can see,it is a lot to do about nothing.

Anonymous said...

It is clear it cannot explain the positive results. The fact that skeptics claim otherwise is par for the course. They have either not understood it, or more likely haven't even bothered reading the article.

But nevertheless I do not agree that it's a lot to do about nothing. It might at least partially explain why skeptical researchers have a propensity to get negative results.

Subconsciously skeptical researchers might want the receiver to fail. Thus if the receiver gives a correct answer which the skeptical researcher also telepathically picks up as being the correct answer, then s/he may have a propensity to question that answer.

And of course a skeptical researcher will suppose the receiver is essentially guessing even though the receiver might feel he's not. So the skeptical researcher may well have a propensity to question the answers given by the receiver anyway.

Anonymous said...

Have the receivers mentations judged by a panel of independent blind judges, which I believe has taken place with positive results.

Post a Comment

Note: only a member of this blog may post a comment.