Dean Radin on Psychic Gorillas
I’ve personally argued with a fellow experimental physicist about the validity of research done at PEAR (Princeton Engineering Anomalies Research Center). I’ll confess that I am unfamiliar with the vast majority of the mathematical and statistical babble in the PEAR papers, but over on the archives of Good Math, Bad Math Mark has some excellent rebuttals. My chief objection to the PEAR studies doesn’t really take a dissection of their methods to get to: their studies are not double blind.
The most frequent counter to my rallying for double blind studies is the claim that the data will be the same regardless of the blindness of the study. This I agree with. However, what will change is the researchers’ interpretation of the data. To clarify the importance of double blinded studies, see the control group article on Skepdic. Personal anecdotes and discussion of the actual subject line occur below this wonderful fold.
To emphasize the importance and relevance of the double blind study, I can relate some experience I had as a wee undergraduate working on the MiniBooNE experiment at Fermilab. In a (very small) nutshell, the MiniBooNE experiment set out to either confirm or deny the results of an earlier experiment called LSND. The LSND results were not compatible with the current picture we have of the standard model of particle physics, which includes three Generations of particles each containing a quark, a charged lepton, and a neutrino. If the LSND results were correct, we would need “new physics” to explain them, one possibility was the existence of a fourth generation of fundamental particles.
The phrase “new physics” is enough to send any physicists heart aflutter. Good science dictates, however, that one filter out any personal bias when analyzing data. MiniBooNE’s solution was to put all their data in “the box.” In this way, the analysis could be worked on, the Monte Carlo simulations could be done, without the researchers either consciously or unconsciously trying to make “new physics” come out of the data. I don’t think it’s too out of line for me to demand the same safeguards used by particle physics’ finest be used when testing psychic phenomena.
Which finally brings me around to the subject line above: Dean Radin’s psychic gorillas. In an article What Gorilla?: Why Some Can’t See Psychic Phenomena, which is also commented on at The Rogues Gallery, Radin makes some grandiose claims about “Big Science.”
That[, the western scientific] worldview, like any set of cultural beliefs inculcated from childhood, acts like the blinders they put on skittish horses to keep them calm. Between the blinders we see with exceptional clarity, but seeing beyond the blinders is not only exceedingly difficult, after a while it’s easy to forget that your vision is restricted.
In this statement I believe Radin might be the only victim of the described restricted vision mentioned in the article. The “blindness” in scientific studies does not blot out reality at all, but the personal biases of those doing the research. By inducing “blindness” researchers only reduce the tendancy to emphasize personal convictions and, in fact, emphasize, you know, reality. By removing the blinders and opening up data to the whimsical interpretation of the researcher negates it’s usefulness in any sort of rigorous discussion of the facts.
Exclusion of these phenomena creates a Catch 22: Human experiences credibly reported throughout history, across all cultures, and at all educational levels, repeatedly tell us that psychic phenomena exist. But Big Science — especially as portrayed in prominent newspapers and popular magazines like Scientific American — says it doesn’t.
I find this statement curious. Because, as Carl Sagan discussed in his classic book The Demon-Haunted World, the interpretation of “mysterious” events changes radically over time. So which is it? Is the phenomenon that modern medicine calls sleep paralysis really a visit from extraterrestrials? Or is it a visit from a demonic Succubus or Incubus as previous centuries’ popular culture would have us believe? What methods would Radin propose for us to determine this?
Really, the fuel of these sorts of interpretations is their inability to be tested. Radin claims of psi phenomena that, “Such effects tend to be small in magnitude, they are highly reactive to the psychosocial context and other environmental factors, and they take substantial amounts of careful data collection to overcome the statistical noise generated by dozens of poorly understood interactive factors.” When you’re dealing with statistics at this level where your noise and your signal are of the same order, you are not in any position to say anything conclusive without numerous checks in place.
Back in the world of particle physics, when dealing with very low signal, high noise sorts of experiments like neutrino oscillation or dark matter searches you rely very heavily on methods to “veto” what is, in all likelihood something you’re not interested in. When you want to study particles that you know will only interact once, partition your detector and reject “hits” that appear in multiple partitions. You’re free to claim that you’ve detected WIMPs based on your noise happening at a slightly higher rate than predicted by your Monte Carlo, but it’s extraordinarily unlikely that anyone will take you seriously. Not because scientists don’t want to detect dark matter, but because it would be bad science.
Tags: Dean Radin, links, statistics