Review of: "EXPERIMENTER EFFECTS AND THE REMOTE DETECTION OF STARING"

[Im a Hedge]

Thanks to Larry Boy for providing this paper to me. The paper is available here: http://www.richardwiseman.com/resources/staring1.pdf

This is a review of the paper EXPERIMENTER EFFECTS AND THE REMOTE DETECTION OF STARING by Richard Wiseman and Marilyn Schlitz. Wiseman is a skeptic of psi and Schlitz is a proponent of psi. Each had previously conducted independent studies on the phenomenon of people detecting when they are being stared at. Wiseman reported negative results and Schlitz reported positive results. This experiment was a collaboration between the two. This is an excellent approach, and I hope to see more such studies.

The basic setup is that participants are isolated in a room with a closed circuit video camera. The experimenter is in another room with the monitor for the camera. A series of trials are conducted, in which the experimenter either stares at the subject on the monitor screen, or does not stare at the subject. The subjects Skin Resistance Level (SRL) is measured throughout the session. The quick version of the results is that subjects had a significantly different SRL when being stared at by Schlitz, compared with their non-staring trials. There was not a significant difference when Wiseman was the starer.

On the face of it, this looks like good evidence in favor of the effect of remote detection of staring. It is also evidence in favor of an experimenter effect. That is, the psi-proponent generated positive results while the psi-skeptic failed to do so. This would be consistent with some proposed models for how such abilities might work. Schlitz may have a greater psi ability level than Wiseman. Also, Schlitz may have a more positive attitude towards the ability, while Wiseman expects that it doesn't work, resulting in a self-fulfilling prophecy. That is, if there is some mental effort, or intent, involved in sending the signal, Wiseman may be failing to do so as a result of his prior belief that he does not have this ability.

One major flaw in the experimental design is (if you've read my other reviews, you know what's coming..) lack of negative controls. Maybe psychologists just don't do negative controls? A negative control would be easy in this design. Have some subjects come in, go through everything just as in the experimental sessions, choose one of the randomized lists of 'stare'/'no stare' trials, but don't have the experimenter go into the monitor room and do any staring. Inconvenience is not a sufficient reason to exclude proper controls, but this wouldn't even be difficult in this experiment. It would be simple.

Aside from that, the experimental design is pretty solid, but could still be improved. The order in which the two experimenters conducted sessions is not randomized. They describe it as 'opportunistic', meaning whichever of the two was available when a subject arrived performed the session with that subject. A proper randomization of this would be an improvement. The experimenter met the subject and described the protocol to them just prior to the experiment session. Using a third party, not one of the experimenters, to do this task would be an improvement. Ideally, this third party would not know which of the experimenters would be conducting the session. This would reduce the ability for the subjects to pick up on any subtle cues about expectations of the experimenters.

Further complicating the protocol, additional experimenters could be used (other than Wiseman and Schlitz). These experimenters would answer questionnaires about their belief in psi, similar to the subjects in this experiment. The level of psi belief could then be correlated with the effect, providing more than just two data points to tests the hypothesis of experimenter effect.

As a final issue, I commend the authors for explicitly addressing the question of fraud. If I remember correctly, Wiseman has experience as a illusionist. Still, it wouldn't hurt to consult another illusionist to be more sure that the potential for fraud has been ruled out. I don't see where there's much potential, given the experimental design, but that's outside of my field of expertise. It has been shown quite dramatically that scientists are not always good at detecting conjuring tricks.

In summary, I find this study to be an impressive example of collaboration between proponents and skeptics of psi. The experimental design is flawed, but not necessarily fatally so. The results indicate some ability for humans to respond to being stared at unknowingly. The results of this study warrant additional investigation into this phenomenon, perhaps with modifications to the protocol along the lines suggested above.


I am a Hedge