Tuesday, August 7, 2007

Parapsychology dispute

Looks like two of George Hansen's initial blog posts amount to swipes at Dean Radin's statistical credibility.

Disputes in academia are nothing new, but thanks to the internet, and the fact that both Radin and Hansen are bloggers, this one has the potential to be a lot more public than many previous dust-ups in parapsychology. . .

UPDATE: Radin has addressed Hansen's critique here. H/T Book Surgeon, in the comments. . .

11 comments:

Anonymous said...

Radin's statistical credibility is all of his credibility. If he cannot rebut Hansen's latest swipe, he is discredited. I wonder, will he even try?

M.C. said...

I disagree.

Applied statistics is one of the most difficult fields to do properly, and Radin is trained in electrical engineering and psychology, not applied statistics.

Certainly it raises very large questions about any of his statistical analyses (if he does not answer this criticism effectively) but it should not completely discredit his research. After all, the raw data can always be re-analyzed if statistical errors are found.

Frankly, I'd like to see science move towards a norm where the raw data is always available for anyone to re-analyze statistically.

Anonymous said...

If Radin wasn't given the requisite statistical training, he never should have become a parapsychologist, let alone the most well-known researcher in the field. But his “Environmental Modulation.." paper's most egregious error is one that no one with his education should have made. If he is that careless and incompetent with the most important facet of his work, it doesn't just call all of his research into question, it puts it into disrepute.

M.C. said...

Anonymous,

As the link I posted above notes, 38% of the papers in Nature, the most prestigious journal in all of science, show statistical errors.

Why are you holding Radin and parapsychology to a much higher standard than the most prestigious science journal? Why do you think that parapsychologists should have better training in statistics than scientists in all other fields?

Ersby said...

Dean Radin's replied to Hansen's first point on his blog:

http://deanradin.blogspot.com/2007/06/two-recommended-books.html

About the second one, I'm not sure what to make of it. Hansen says the figures he gets are still stastically significant, but he hasn't applied the Bonferroni correction (NB, I'm not trained in statistics either - take what I say with a pinch of salt). If you do that, then neither Radin's nor Hansen's findings are significant.

But then, as Hansen said, the different correlations aren't independent, so the Bonferroni correction shouldn't be used. So... what?

His last point, about pondering the implications of Schlitz workig at the Neotic Institute goes over my head somewhat.

Tor said...

I feel that the easiest way to resolve this would be for Hansen to contact Radin directly to clear things up.

If Hansen is insinuating that Radins work at IONS is sloppy because he is working under Schlitz (wich was in charge of reviewing the papers, including the paper in question, at the conference he mentioned), then I think he is far off.

Radin's papers are published in peer reviewed journals, which should show that his work has the minimum level of quality required 8taking into account MC's remarks). That he is working at IONS, under Schlitz, is irrelevant.

Anonymous said...

M.C.,

The issue isn't Radin's lack of statistical training - he's had ample. Enough, probably, to justify his pretense of being an expert. And more than enough, surely, to have prevented the blunders Hansen noted. Moreover, as a past president of the Parapsychological Association and the field's most visible representative, Radin has to be held to a high standard. He has a responsibility to be a competent scientist. If a president of the American Psychological Association were exposed in a similar fashion, it would cause a scandal. Radin and parapsychology deserve nothing less.

M.C. said...

Anonymous,

You still haven't answered why Radin should be held to a higher standard than the most prestigious science publication in the world.

Anonymous said...

Is it your stance, M.C., that since Nature's peer-review process is not foolproof, incompetent research should be acceptable? Since bad-quality studies have been published in science's most prestigious journal, a careless scientist like Radin should be given a break, right? In your view, it seems, it is far too much to ask that scientists do their jobs well. Needless to say, I disagree. And as I said before, I merely think Radin should be held to the same high standard as his analogues in mainstream psychology.

Tor said...

The fact that scientists (regardless of field) can make a few mistakes in their career should not come as a surprise. After all, scientists are human, and as we all know, it is human to make mistakes.

On the other hand, if a scientist makes many mistakes, that should not be tolerated as it shows signs of incompetence.

But my point is that the peer review process itself gives a certain amount of quality control.You can't do bad science and expect to get your papers published.

Dean Radin has published a fair amount of peer reviewed articles. This doesn't mean that he doesn't make mistakes (I haven't read the article in question, so I can't comment on it), but I think this shows that he is not sloppy.

Also, in science journals (and science in general) the question of methodology, both experimental and theoretical, is discussed and critiqued all the time. This is how science refines itself.

Book Surgeon said...

Dr. Radin has addressed Hansen's concerns on his blog with a post new today.