« Skeptics Circle | Main | Astrology Still Fails »

July 19, 2009

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

That was great! Please make more.

Don't forget, the thing that gives athletes and movie stars their abilities.

Yup, because letting your team win the pennant is more important than healing amputees.

Skeptico, why do you close comments on your threads? You should leave them open. Especially on so-called "classic skeptico posts". Aren't these issues you address timeless? There should be a way for new material to be added if it is relevant.

In particular, I came across your discussion of Carlson's famous experiment "refuting" astrology. But just recently, a paper has been published by an apparently reputable statistician arguing that his experiment is flawed and in fact can easily be read to support positive astrological hypotheses. But I cannot add a reference to this in the appropriate comment thread, so someone reading your blog entry will not thereby find out about this critique of Carlson.

When will you be addressing this apparently powerful critique yourself?

Benson Bear:

Link to this paper?

I would assume he's referring to this:
http://www.theoryofastrology.com/ertel_carlson.htm

Jeremy is correct. The contents of current issues of the "Journal of Scientific Exploration" however are not available online, and I suspect most libraries do not hold a subscription to it.

I believe you can request a copy of the paper from Ertel himself. Ertel appears to know his stats, certainly better than Carlson did at the time. But is he just an apologist for astrology trying to throw up dust? Or just a troublmaker? I lack the expertise to determine the issue (and so, I suspect, do you).

There is another recent positive statistical test for some astrological type claims in that Journal, done with dogs and their personality types at birth, showing correlations to "angular" objects in the heavens. Again, the person looks like they know their stats, but also, prima facie it looks a little suspect to me, because they just look for every possible combination of influence, and claim a positive result when they find just one or two. They claim they have altered the p values appropriately for this but I can't evaluate if they have. Also, there could be a problem with blinding in this experiment. Anyway, this paper is available online. I'll look for the link later and post it if no-one else did.

Perhaps start a new post since this is a thing of yours. "New positive tests of astrology"...

The paper about the dog experiment can be downloaded
from the JSE site in pdf form. The title to search for on that page is "An Empirical Study of Some Astrological Factors in Relation to Dog Behaviour Differences by Statistical Analysis and Compared with Human Characteristics".

Benson,

You are already suggesting Skeptico should post on "New positive tests of astrology". You're jumping the gun a bit there.

The article Jeremy linked to says that Ertel criticised some aspects of the Carlson study, and "reiniterpreted" the results in a way that offered a slight improvement on the results. Without details of how he did it, we just have to take his word.

Ertel is a parapsychologist, astrologer, and fan of Dean Radin's work. Not promising. You need to get hold of that article, and report its results clearly. Don't expect anyone else to waste a minute of their lives on Ertel's nonsense, just on the sketchy assertions in an that article. If you make an assertion, you need to back it up yourself.

The article about dogs that you linked to instead, as far as I can understand, is actually a clearly negative result for astrology. It's also atrociously designed.

Six behavioural traits were correlated with ten astrological traits. All they found was some kind of "strong"association between the angle of Jupiter and the sun, and extroversion. The other measures failed. Of course, they don't put it like that in the article.

Of course, no control group. They just gather up a bunch of extremely random and subjective data and then interpret away.

I'm afraid, Benson, this is a terrible link to provide for a "positive test of astrology", because astrology is not being tested. For that it would have needed a control group, to gather a comparison sample for chance. All that was being tested was the astrologers' ability to connect two sets of information, with no standard for success or failure set up.

Yakaru,

You would rather have this discussion as off-topic comments in a different thread? Makes sense to post a new article about the recent reinterpretation of a well-known figure in this debate of an extremely famous (even as featured by Skeptico himself) test of some astrological claims.

I got Ertel's paper. Ertel seems as technically competent as many. So rather than calling him names the paper should be addressed, especially since many websites already are referring to it positively (in the form of quoting the same "press release" mostly).

Putting "failed" in boldface doesn't serve any purpose, does it? The way they put it in the paper, as I understand it, is that they said at the beginning they would do all the tests, and then adjusted the p-values required so that getting one or two positive results amongst a lot of failures, would still be a statistically significant result. This does sound prima facie fishy to non-experts in statistics, but they do explicitly address it. Although I suspect in the end this is where the problem lies, in this experiment, along with the dubious blinding. But I said this already, above, didn't I? I don't see why a "control group" is needed, either. There is not a treatment being administered here. They are just looking for a statistically significant correlation.

Nor was the ability of any astrologers being tested.

Another off topic suggestion Skeptico. Two suggestions. Second suggestion: Maybe an open thread where people can post general concerns you might want to take up based on older posts that have comments closed. But first suggestion: Move the excellent list of "fallacies" repeated below from a subsection in the "comment guidelines" to a more prominent location of its own. It's a very good list.

The appeal to be open-minded
The appeal to other ways of knowing
The appeal to “science doesn’t know everything”
The appeal to “science was wrong before”
Ad Hominem
Equivocation
Appeals to Quantum Mechanics
Appeals to personal experience

Benson,

The posts explaining the fallacies you are referring to can be found under "Categories" --> "Fallacies" at the top of the page.

Martin,

Not really, that is not a list of fallacies but apparently posts that discuss specific instances in many cases. For example the second post there is titled "Dinesh D'Souza is Not Very Bright".

I was referring to the section Skeptico has (imo, buried too much) "Skeptico Info" --> "Comment Guidelines" --> "Fallacies". This is (mostly, imo) a very good section that deserves to be more evident. It is well done because it has also in introductory level giving a short paragraph to each one, and then a link to a post in many cases that goes into more detail.

I think that many of the phenomena described there go beyond just episodic employment of specific fallacies as traditionally construed though. The "other ways of knowing" and the "open minded" ones, for particularly good examples, are more of kind of a general sort of mind-set that some people have. They differ from typical fallacies in that they are even happily endorsed. You don't see many people saying, for example "I like to beg the question" or "I like to affirm the consequent".

Arghh, meant to hit preview and then fix some bad phrasing there, but hit post instead. Oh well.

The "other ways of knowing" and the "open minded" ones, for particularly good examples, are more of kind of a general sort of mind-set that some people have. They differ from typical fallacies in that they are even happily endorsed.

appeal to popularity :)

OK, a couple of things.

First - I'm underwhelmed by that description of Ertel's "paper" in what is obviously an extreme woo journal. (Peer reviewed? Give me a break.) I'll go into more detail in a full post, and perhaps after I get a copy of the actual paper (or perhaps not), but you can telll most of what you need by the way it ends:

The results are regarded as insufficient to deem astrology as empirically verified,” Ertel warns, “but they are sufficient to regard Carlson’s negative verdict on astrology as untenable.”

- which is an epic FAIL. If the results of Ertel's study are insufficient to deem astrology as empirically verified, then astrology failed the test - period. The null hypothesis is that astrology doesn't work. Come on, this is basic experimental design. Anyway, I have been in communication with Shawn Carlson this morning, and will write more in good time.

Second - regarding keeping comments open. I have found that after a few months, we get nothing new, just the same lame woo responses that have already been answered. I get tired of reading the same stuff over and over and over again, and of having to give the same rebuttals over and over and over again. The classic posts are the worst in this regard - "What The Bleep" and "The Secret" fans keep posting the same old same old no matter how many times they are introduced to reality. It gets tiring and boring after a while, and this blog is not set up as a playground for woos to promote their stupidity for ever and ever. Life's too short. So after six months (or a little longer) I close posts to comments. If anyone has anything new to offer they can email me. IIRC, on only one other occasion (and that was on astrology too, funnily enough) did anyone have anything new to say.

Third - thanks for the suggestion about the fallacies. Perhaps I should split up the "comment guidelines" page into two - page one for the "rules" and the second page for teh fallacies list. What do people think - would that be a good idea?

I do think a prominent link to display your posts about fallacies would be a good idea. It's worked out reasonably well for me to have my Doggerel Index as a central location.

...I really need to get back to expanding it. Had a few ideas for new posts, but I didn't write them down. :P

Skeptico, also if possible, I think it would be good to think of something else to call them besides fallacies. They are fallacies, but as I pointed out, the really interesting thing about some of them is that they are worldview components that are positively endorsed by those holding them. They prima facie "sound good" and even are, but are easily and often carried into grotesque and erroneous excess.

As for Ertel's paper, after further examination, it does seem like he is grasping at straws. For one example, it's not "unfair" to have a choice from three items, especially since the astrologers agreed to that, even if using a choice from two items is better in that it has more statistical power.

However, it *is* a little unfair, I think, to emphasize that the astrologers thought they would do better than they did. That is not part of the hypothesis being tested statistically, and besides, their erroneous prediction in this regard has two possible component sources of error, only one being overweening confidence in their (presumably non-existent) astrological powers. The second is that they no doubt made rough assumptions about how dissimilar the profiles would be. After seeing them they apparently claimed, and it doesn't seem totally implausible, that they were wrong in this assumption, but that is no reflection upon their astrological skill being tested.

Also, I don't think one can blame Ertel for not having his paper published in a better journal. What journal is going to publish papers like that even if they are right? And it is true that the publication in Nature was in a commentary section, but that this is not always noted.

Anwyay, this comment thread is not the place to go on and on about this. I'd also like to hear comments on the other paper I mentioned, though (same flaky journal). I don't think the study design is "atrocious" as Yakaru says. It seems a standard legitimate way of "data mining" a data set with the tests for significance appropriately altered given the fact that only a few of the many tested correlations statistically deviate from chance. The problem appears to be with blinding, which is not even mentioned. Also possibly with the pre-test selection of the astrological aspects being tested. Why were these apparently strange aspects ("angular" or not), chosen? If there was a second level of searching through possible data sets, this one not mentioned in the paper, in order to come up with this set of aspects which then contained a few significant correlations with dog personality factors, then *that* could even be worse than "atrocious design", indeed a form of fraud. But I doubt that was actually done. Blinding seems to be the worst problem here.

The principal investigator in this study is reported as having had a "journey to Damascus". This is another kind of "fallacy" one might add to the list that one often hears in these contexts.

I think it would be good to think of something else to call them besides fallacies. They are fallacies, but as I pointed out, the really interesting thing about some of them is that they are worldview components that are positively endorsed by those holding them

sorry if I'm not polite because this was not directed to me but to Skeptico, but I guess my first reply to this concept was not clear enough, I just said "appeal to popularity", but maybe I have to explain it further:
just because it sounds good to some people and they use it as a "worldview", it doesn't mean it's automatically right or not a fallacy. I think it would be some kind of "political correctness" to call them other things instead of logical fallacies, but, sorry, the truth or not of something is not supposed to be politically correct.
if somebody supports astrology or new age (rhymes with sewage) stuff saying "you have to be more open minded!", we should not be forced to accept it "ok...that's just your worldview, it's ok", it smells like the kind of undeserved respect religion try to impose for itself on society.
and in fact, they are the close minded ones because they are not open to the idea that they might be wrong about astrology, new age, whatever.

it doesn't matter if a lot of people think it sounds good, it's not valid and the popularity of the idea doesn't make it so.

I hope I was clear enough now,
good bye!.

Pelger,

Yes, for me you have to explain further, since I could make no sense out of what you were saying. As I suspected, you completely misunderstood what I was saying. Of course, just because something sounds good to someone, or to many, and just because it is a "worldview", doesn't make it right or "not a fallacy". My point was that these two phenomena in particular--appeal to "open mindedness" and appeal to "other ways of knowing"--are in some sense even *worse* that what are traditionally called fallacies, precisely *because* they are (erroneously) seen as "good" by the people who are engaging in these appeals. No one is going to proudly stand up and say "I've just begged the question", but they will stand up and proudly say that they are "open-minded" or endorse "other ways of knowing". That doesn't make what they are doing right, as I think should be evident. But in terms of understanding the phenomena, and trying to curtail it, I think it is important to see the difference between this and committing a less emotionally loaded form of fallacy.

Skeptico,

Incidentally, looking back at your original discussion of Carlson, I see there are some errors in the description you give of his experiments.

First, you say that the subjects were given three "astrological charts" and asked to say which of them were theirs. They were not given charts. They were given interpretations of the charts in totally non-astrological, psychological terms. And their failure here did not prove anything, as Carlson observed, since they were also unable to pick out their own "scientific" CPI profiles.

Second, you say the astrologers were given "natal data (date, time and place of birth)". But they were not. They were given natal charts. So at least they were spared making errors in chart preparation.


ok, I think I got what you say, you see "appeal to other ways of knowing" and "be open minded" as some kind of social phenomenom, right?.
but I don't see why it disqualifies them as just plain logical fallacies like "straw man" or "appeal to authority", the fact that some people endorse them happilly as if it was some kind of virtue, could be described as a side effect or even as part of the definition of the logical fallacy, as in "the woo that appeals to be open minded, usually thinks it's a good idea so he tends to skip the filter of reason and critical thinking, and tends to be guillible".

Sheesh, you're previous post was bad enough. I get so tired of anti-theists trying to shove their beliefs down people's throats just like theists do. Get over yourself. If you're content believing what you believe, great. If somebody else is content believing the opposite, great. Anti-theists have done horrible things, theists have done horrible things. blah blah blah. You're all so smart. Congratulations on having figured It all out. Both world views believe something came from nothing. The rest of us don't care.

Riiiiight. Posting his thoughts on his personal blog is shoving it down your throat because you have no choice about whether or not you read it.

The rest of us don't care.

Clearly false. You care, or you wouldn’t have bothered posting.

Foo Fighter:

I get so tired of anti-theists trying to shove their beliefs down people's throats just like theists do.

Who made you read this post?

Get over yourself.

Indeed.

The comments to this entry are closed.

Search site