Roberts Matthews and Daniel Adkins discuss how sceptical scientists should be when faced with the implausible.
ESP article elicits protest
Last week regular Frontiers contributor Robert Matthews discussed a journal paper purporting to provide evidence for extra-sensory perception. That led to a lively exchange of e-mails with Daniel Adkins, the academic director of Murdoch University in Dubai. This is their discussion:
I am shocked that a person of your stature would put your name to an unscientific article.
First, if we allow that a statistically significant result was found, "Experts" are correct to withhold provisional assent to ESP as one experiment of any kind is insufficient to establish a theory in science. The headline and content are therefore extremely misleading in implying that evidence exist which is sufficient to say that those who do not agree are in denial.
Second, given the plethora of experiments that have been conducted on ESP, where every credible experiment has shown no statistically meaningful support for ESP, one anomalous experiment should lead the scientists to question their own protocols before making a public statement regarding their results.
It should also cause the scientists to abandon the freshman level "p-values" [probability values] and use Bayesian analysis instead. Given the prior probability as shown in the other well-conducted experiments, a much more significant result would be necessary to find that a credible result overturning years of other experiments had been found.
Third, if ESP existed then the Million Dollar Challenge from the James Randi Educational Foundation would have been claimed years ago. To this day, no one under mutually agreeable conditions has been able to demonstrate any form of ESP. It is fairly clear that if this phenomenon existed someone would have spent the few hours necessary to claim $1m.
What you have done brings disrepute upon yourself and intellectually honest scientists everywhere. It also makes my job as an education professional more difficult as I have to explain to students why this is not valid science as presented.
Even worse, it distorts the views of the general public in regard to what is real and what real science can demonstrate.
Most respectfully your colleague,
Dear Mr Adkins
Many thanks for your response. I'm somewhat perplexed by your comments. Nowhere in the piece do I say that precognition has been proved; the headline simply states that the research has reached legitimacy, while the final sentence of the piece explicitly states that more evidence is needed.
The piece as a whole is simply a statement of my belief that more scientists should join in the process of either confirming or refuting Prof Bem's findings, as the ultimate proof lies in the application of the scientific method, rather than (yet more) armchair critiques.
I certainly agree with you that Prof Bem's research should be a call to move away from p-values; I've long been an advocate of Bayesian methods and have published much on this in both refereed journals and the media, and have explicitly called for the use of Bayesian methods in parapsychology.
Indeed, his papers have sparked a big debate about precisely this, and I hope it kindles more interest in this approach.
All best wishes
Thank you for your reply. I do believe that you are being somewhat disingenuous. If you truly support using Bayesian methods, then given the prior probabilities, this study would have moved the probability such a small amount as to be nearly negligible.
Given that, to have an article titled "Experts Deny ESP Despite Evidence" is clearly irresponsible hyperbole. The phrase "deny despite evidence" would be appropriate to apply to those who do not acknowledge the strength of scientific theories that have withstood numerous well-designed challenges, such as general relativity.
Basically all credible scientific evidence demonstrates that ESP does not work. When challenging the established science, to claim a single study makes all other experts "deniers" is so wrong as to offend intellectual honesty.
Rouder and Morey had the scientifically accurate take on Bem's study and what you wrote does not present the significance of the study fairly.
You also did not address the fact that anyone who could demonstrate ESP has had over a decade to claim a million-dollar prize for a few hours' effort, yet despite many claimants, no one has been able to demonstrate ESP.
Most respectfully your colleague,
Here's what Rouder and Morey say:
"We develop a meta-analytic Bayes factor that describes how researchers should update their prior beliefs about the odds of hypotheses in light of data across several experiments. We find that the evidence that people can feel the future with neutral and erotic stimuli to be slight.
"There is some evidence, however, for the hypothesis that people can feel the future with emotionally-valenced nonerotic stimuli. Though this value is certainly noteworthy, we believe it is orders of magnitude lower than what is required to overcome appropriate skepticism of ESP."
... or, to sum it up, "Experts say there is evidence for ESP, but they don't find it compelling". I'd agree with that entirely - and it's also pretty much exactly what the headline said.
I do hope you are saying this tongue-in-cheek. The word evidence as used below is in the scientific manner meaning "an observation has occurred" not the manner in which it would be taken by most of the public and which would be necessary to justify the headline, specifically the "CSI" definition which the public would take as "evidence that would stand up in court".
When Rouder and Morey say that the "evidence" is "orders of magnitude lower than what is required to overcome appropriate scepticism of ESP", they are saying exactly what I am saying; a well established scientific understanding is not overcome by one anomalous observation which has failed to be reproduced despite five documented attempts.
You know as well as I do that it is a logical fallacy to take a word that has one specific meaning as used and use it with a different meaning to make a case. You are doing this with the word evidence, where it has a scientific meaning in the article and a common meaning as interpreted by the general public in the headline.
All the best,