
View previous topic :: View next topic 
Author 
Message 
Tom Moderator
Joined: 11 Oct 2003 Posts: 3300 Location: New Jersey, USA

Posted: Sat Nov 27, 2010 7:39 pm Post subject: Statistics 


There is no astrology (that I know of) in this article. Rather it is a challenge to the omnipotence of statistics in the minds of scientists. I have not read all of it, and what I have read I admit I skimmed, but there may be something of interest that is applicable to astrological studies. I hope so.
http://www.sciencenews.org/view/feature/id/57091/title/Odds_Are,_Its_Wrong


Back to top 


Ed F
Joined: 22 Jan 2008 Posts: 301 Location: Ipswich, MA USA

Posted: Sun Nov 28, 2010 3:46 pm Post subject: 


Nice one, Tom. Thanks! 

Back to top 


mattG
Joined: 21 Sep 2007 Posts: 345 Location: Greenwich UK

Posted: Mon Nov 29, 2010 12:30 pm Post subject: 


Yes, as you say, no astrology but interesting nonetheless. The article relates to medicine mainly but I remember a member of parliament over here saying "We must have the science" What he meant was that no one could make a decision about legislation without statistics.
Before I consider the implications for astrology I must suspect that we pay public servants a lot of money to produce bad data so politicians can make stupid laws that we do not need.
O for wise rule
Matt 

Back to top 


yuzuru
Joined: 01 Apr 2005 Posts: 1360

Posted: Mon Nov 29, 2010 12:53 pm Post subject: 


I had written my view on the matter, but I think skyscript ate my homework...
So, thanks for the link. It is an important point, but few people knows it. In fact, there are a lot of other statistica problems that were not discussed in the article. _________________ Meu blog de astrologia (em portugues) http://yuzuru.wordpress.com
My blog of astrology (in english) http://episthemologie.wordpress.com 

Back to top 


GarryP Moderator
Joined: 23 Oct 2003 Posts: 208 Location: UK

Posted: Tue Nov 30, 2010 4:44 pm Post subject: 


Thanks Tom, top man! I've been busily sharing this link around.
Cheers,
Garry 

Back to top 


Ed F
Joined: 22 Jan 2008 Posts: 301 Location: Ipswich, MA USA

Posted: Tue Nov 30, 2010 6:46 pm Post subject: 


Quote:  But Bayesian methods introduce a confusion into the actual meaning of the mathematical concept of “probability” in the real world. Standard or “frequentist” statistics treat probabilities as objective realities; Bayesians treat probabilities as “degrees of belief” based in part on a personal assessment or subjective decision about what to include in the calculation. That’s a tough placebo to swallow for scientists wedded to the “objective” ideal of standard statistics. “Subjective prior beliefs are anathema to the frequentist, who relies instead on a series of ad hoc algorithms that maintain the facade of scientific objectivity,” Diamond and Kaul wrote. 
Now, as a thought experiment, consider astrology as "being like" statistics. Instead of probabilities, it gives "indications" that can be contextualized, interpreted and applied in many ways. Is astrology really so different from applied mathematics?
 Ed 

Back to top 


Olivia
Joined: 15 Oct 2008 Posts: 866

Posted: Tue Nov 30, 2010 7:50 pm Post subject: 


Probably. How many times do we use one set of personallytailored statistics to predict what one individual will do?
I find it hard to believe that everyone gets their own statistical data  most of it is of the aggregate variety and tends to predict that if you're 1834 and male, you will buy....
Just tell me that drug companies, car companies, and whatever, aren't electing times for testing to prove that their product is efficacious/safe. That could probably be manipulated quite easily. But I don't hope to live to see it. 

Back to top 


Eddy
Joined: 04 Feb 2009 Posts: 922 Location: Netherlands

Posted: Tue Nov 30, 2010 8:45 pm Post subject: 


Olivia wrote:  I find it hard to believe that everyone gets their own statistical data  most of it is of the aggregate variety and tends to predict that if you're 1834 and male, you will buy....

Several things similar to this are already happening nowadays. Health insurance companies estimate the life length and statistical chances for diseases of their clients to calculate the amount of health insurance to be paid by them and finally to lead to profit for the company. 

Back to top 


###
Joined: 08 Jul 2004 Posts: 1381

Posted: Tue Nov 30, 2010 9:26 pm Post subject: 


The field of statistics grew as astrology faded away in the late 17th and 18th centuries. Astrology nearly died in the 18th century while statistics was taking off. A new philosophical view moved in. What killed one area of study fed the other – possibly a good indication that astrology and statistics don't belong together.
This may sound trite and maybe is said too often, but literature, music, poetry religion and flights of imagination gain little through statistical analysis. It's an odd person who applies statistics to those activities. Astrology does have a large element of intellectual method, but ultimately the insight comes from imagery (and the key here is nonmeasurable insight). Each chart is a composition of imagery. Application of statistical method kills the living connection to imagery. It kills living insight. 

Back to top 


Ed F
Joined: 22 Jan 2008 Posts: 301 Location: Ipswich, MA USA

Posted: Tue Nov 30, 2010 9:35 pm Post subject: 


My meaning was actually more general than about just statistics; the comments so far have focused on the particular nature of the indications that statistics give. What I proposed for consideration is the similarities between the application of mathematical and astrological models given the comments about the limitations of statistics that were presented in the article.
We seem to have an order of abstraction issue so far in the discussion, which is perhaps my fault since I seem to have pressed some ready and waiting buttons.
Consider the practitioners' activities involved in the use of each kind of model.
 Ed 

Back to top 


Eddy
Joined: 04 Feb 2009 Posts: 922 Location: Netherlands

Posted: Wed Dec 01, 2010 10:42 am Post subject: 


Rather than pointing out the shortcomings of statistics the article points out the incorrect use of statistics in science, a problem that professional statisticians themselves have been warning for.
the article wrote:  How could so many studies be wrong? Because their conclusions relied on “statistical significance,” a concept at the heart of the mathematical analysis of modern scientific experiments. 
& wrote:  Another common error equates statistical significance to “significance” in the ordinary use of the word. Because of the way statistical formulas work, a study with a very large sample can detect “statistical significance” for a small effect that is meaningless in practical terms. 
This was also concluded from an observation of the only Gauquelin study that (for astrology) yielded positive results. http://www.astrologyandscience.com/ghist2.htm under § "Use of effect sizes"
Geoffrey Dean wrote:  Effect sizes are usually expressed as a correlation where 0 = no effect, as between one coin toss and another, and 1 = perfect correlation. Thus if all soldiers had the same sun sign, or if all children had the same aspect as their parents, the effect size in each case would be 1. An effect size involves the whole sample and thus gives a good idea of what is happening, whereas looking at just part of the sample can be misleading. Thus the typical planetary surplus or deficit in key sectors of 10% to 25% may seem like a lot, but the corresponding effect size (which for 40 years nobody bothered to calculate) is only 0.02 to 0.05. Planetary effects are tinier than they might seem. 
& wrote:  Unfortunately Gauquelin never used effect sizes. Instead he generally presented his findings in terms of statistical significance, which varies roughly as N2 where N = sample size. As N increases, even the most trivial effect will eventually reach astonishing significance. 
I'll give a 'gauquelianlike' example: If we want to test the significant house position of Mars from charts of 1200 sports champions, then according to chance 100 champions will have Mars in I, 100 in II and so on. This is a chance of 8,33%. If studies would point out that 122 champions have Mars in I and 98 champions in each of the other houses then the statistical significance of Mars in I is the relation of 122 to the chance value of 100. The statistical significance or the chance that a sports champion has Mars in I will be 22%. This sounds quite impressive but in terns of distribution over the houses this means that 122/1200= 10,17% of the sports champions have Mars in I and 98/1200= 8,17% of the sports champions will have Mars in each of the other houses. This is hardly a significant difference. Although there's an effect it doesn't mean a lot in practical use.
Tom Siegfried concludes his article with the suggestion of the Bayesian approach of the use of previous knowledge included in the statistics. His example ('Box 4: Bayesian Reasoning') on the last page of the article about the drug tests of steroids users in baseball is interesting. The question is how this could be applied in astrology. First the premise will be problematic. In the example a premise of 95% of the accuracy of the drug test is used. Let's imagine other accuracy premises.
 In an extreme way one could imagine a 100% accuracy if the to be tested baseball players are never left alone or permanently left under camera survey. In case of a 100% accuracy the chance that a positive test is really correct is also correct is also 100%. This is alwasy the case; the prior probability will never make any difference.
 One could also consider that the chance that the test is correct is 50%, or the chance of tossing a coin. In this case the chance that a positive test is correct will always be equal to the used prior probability.
 In another extreme way one could think of a 5% accuracy. Imagine that the premise is that everybody whose name is John uses steroids is accurate for 5%. A ridiculous example perhaps but when applied to the example in the article it will lead to the conclusion that the chance that baseball players who are called John use steroids is 1 : 361, fortunately very small.
What with astrology then? Which premise should be used. No book on astrology ever states that the chance that a planet X in position Y means Z is ...%. The indications are usually stated in a way that the accuracy is 100%. Still this leads to the problem how we came to that idea in the first place. The steroid drug test example uses a 95% accuracy. However in the first place som laboratory tests must have been made to develop this test method, for example by using certain substances that will colour blue when exposed to a sample of steroid drugs. The lesser accuracy than laboratory circumstances which is due to the digestion in the human body will lead to the premise that the test will be somewhat less accurate. Anyhow, some previous testing must have taken place to come to the premise in the first place. If this previous (laboratory) research never took place, the premise will be based on nothing and won't provide better than chance or even worse as in the 'John' example.
No previous tests of the astrology characteristics of planets are known of; the meanings of planets and signs were mainly based upon anology and later became generally accepted as fixed rules. However these astrological rules are considered as 100% accurate. The good news is that because of this astrology can be tested through statistics. The bad news is that all tests turned out to be negative.
There are three main possible reactions to these negative results.
 one can conclude that astrology is incorrect and quit the practice;
 one can conclude that statistics are an improper way to test astrology, or;
 one can conclude that characteristics of planets/signs can't or should not be set in rules.
*Apart from a possible personal drama of disenchantment, the first reaction is not a real practical problem.
*The second reaction is a bit problematic in my opinion be cause I tend to believe in statistics. If astrological features are presented as facts then they can be tested. Complete denial could lead to an ostrich attitude. Most often the denial is based upon the meaning that is ascribed to the great merit of astrology. However this invariably leads to the old objective versus subjective discussion or fact versus meaning. The problem with this is that people will be discussing different things and the discussion never ends. Another problem is that different schools can ascribe different meanings to astrological issues, tropical versus sidereal for example. The disbelief in testability leads to the problem that two totally opposite systems can coexist (e.g. Saturn as a benefic and Saturn as a malefic), which is illogic.
*The last reaction could be that all dogma's and schools should be abandoned and just a handful somewhat flexible criteria could be used as premises. Not the premise that planet X in position Y definitely means Z woud count but rather the premis that planet X in position Y could mean A in this individual or B in the other or C under certain circumstances. The problem of this view is that it necessarily leads to vagueness and testing astrology will be very difficult, it becomes nonfalsifiable, and in a far advanced stage astrological technique will become useless from a practical point of view. 

Back to top 


Richard Vetter
Joined: 03 May 2008 Posts: 54 Location: Offenburg, Germany

Posted: Wed Dec 01, 2010 1:08 pm Post subject: 


Kirk wrote:  The field of statistics grew as astrology faded away in the late 17th and 18th centuries. Astrology nearly died in the 18th century... 
Man lost his connection with the powers of fortune and destiny. He didn't feel ruled by the supernatural any more. Instead, he thought himself to be the maker (manager) of his life. God and sensefulness was replaced by pure  and manipulable  chance. _________________ http://astroinfo.astrologix.de/english.htm 

Back to top 


waybread
Joined: 05 Mar 2009 Posts: 629 Location: Canada

Posted: Wed Dec 01, 2010 11:47 pm Post subject: 


I liked this article a lot. Some of it seems like common sense, even for quantificationphobes like myself. I think it does have relevance for astrology.
For example, oftentimes scientific studies cannot be safely generalized beyond the limits of the study parameters, yet this happens all the time. For example, a medical study might show that young white males eating guavas show a statistically significant drop in harmful blood cholesterol levels. Yet when you read the study, actually the researchers looked at only a very small sample of men (like maybe 2 dozen) for a very short period of time (like maybe 2 weeks.) Nonetheless, we're out there buying guavas, in the hope of a miracle cure!
Another problem, is that a statistical study itself only shows a correlation between variables at a level higher than chance. If two variables constantly covaried by the identical amount, you'd have a candidate for a scientific law, not a hypothesis in a statistical test. You normally get all kinds of data points that don't fit the curve or model. Yet the study design really cannot explain the data that don't fit snugly into the correlation. The scientist might attempt to do so, but only by conjecture or referal to other studies that may or may not be a good match with her research. In the above hypothetical example, it would not explain why some of the men in the guava study did not show a drop in triglyceride levels. This is similar to the "wellfed, barking dog" in the article's example.
Some of us have talked a bit about Geoffrey Cornelius, The Moment of Astrology. He argues against widespread statistical studies, partially on the grounds that some big studies in the past showed no significant results. A major example was a study of suicides in New York, where neither statistical tests of horoscope placements, nor astrologers' delineations, produced results beyond random chance. He makes the point that a horoscope, like a life, is a complex and highly individualistic thing.
So this is why I think one might possibly get astrology to the point where a really large, sophisticated statistical study of horoscopes might produce meaningful results. But then the problem would be, how do you explain the outliers in the data? If 80% of the data behave as expected, the study is probably still no good in explaining why the remaining 20% got out of bounds, even where the 80% correlation is highly significant.
This might not be such a problem for simple personality delineation. But it would get huge with crunchtime deathclock predictions or financial adivising in astrology.
To quote the author:“Determining the best treatment for a particular patient is fundamentally different from determining which treatment is best on average,” physicians David Kent and Rodney Hayward wrote in American Scientist in 2007. “Reporting a single number gives the misleading impression that the treatmenteffect is a property of the drug rather than of the interaction between the drug and the complex riskbenefit profile of a particular group of patients.”
This is another good argument for astrology's focus on the individual horoscope. To put it differently I might suggest that a woman with Venus in Scorpio exhibits jealousyexcept when she doesn't. And when and why she doesn't should be indicated by additional horoscope factors that (hopefully) I am smart enough to detect. Maybe the computer can do a better job of this than I could, but so far, so bad.
Insofar as I understand the "Bayesian" approach to statistics, this wouldn't fare much better, as it opens the door to a lot of pretrial biases.
This is why one needs a lot of studies to affirm or reject a "scientific fact." A single study can be prone to all kinds of design problems. 

Back to top 


waybread
Joined: 05 Mar 2009 Posts: 629 Location: Canada

Posted: Wed Dec 01, 2010 11:55 pm Post subject: 


Olivia wrote:  Probably. How many times do we use one set of personallytailored statistics to predict what one individual will do?
I find it hard to believe that everyone gets their own statistical data  most of it is of the aggregate variety and tends to predict that if you're 1834 and male, you will buy....
Just tell me that drug companies, car companies, and whatever, aren't electing times for testing to prove that their product is efficacious/safe. That could probably be manipulated quite easily. But I don't hope to live to see it. 
It's here, Olivia. it is called commercial geodemographics, or geodemographic segmentation. I don't know that the consultants use astrology, but that could be next!
Eddy, thanks for your clarity! I would like to propose a 4th option for the "failure" (if it is that) for statistical studies to date , to support astrology. Maybe the studies could have been much better designed. For example, in the NY suicide study (which I've not read), I believe the researchers looked only at birth charts, not at transits or progressions. Also, I have looked up some French birth records on Astrodienst, as including individuals surveyed in the Gauquelin research. Holy moly. Prior to the early/mid 20th century, most of the times were rounded to the nearest hour. 

Back to top 


Eddy
Joined: 04 Feb 2009 Posts: 922 Location: Netherlands

Posted: Thu Dec 02, 2010 3:36 pm Post subject: 


waybread wrote:  I would like to propose a 4th option for the "failure" (if it is that) for statistical studies to date , to support astrology. Maybe the studies could have been much better designed.  You're right, my list wasn't meant to be limitative by the way. Yesterday I had a look again in the article because I hadn't read the 'Boxes' 1  3. 'Box 3: Randomness and Clinical Trials', could imply what you propose.
the article wrote:  Understanding the individual differences affecting response to treatment is a major goal of scientists pursuing “personalized medicine,” in which therapies are tailored to each person’s particular biology. But the limits of statistical methods in drawing conclusions about subgroups of patients pose a challenge to achieving that goal. 
Perhaps this could be incorporated in the 'choose the right chart for the description of this and that person tests'. Then not only the chart data should be given but also social/cultural/economic background. I have the impression that in astrology books this isn't done a lot in a systematical way, but in daily practice it is done whether unconsciously or not. Ptolemy already wrote about these background issues.
Tetr. IV10 Tetrabiblos wrote:  So then, as, among all genethialogical inquiries what. so ever, a more general destiny takes precedence of all particular considerations, namely, that of country of birth, to which the major details of a geniture are naturally subordinate, such as the topics of the form of the body, the character of the soul and the variations of manners and customs, it is also necessary that he who makes his inquiry naturally should always hold first to the primary and more authoritative cause, lest, misled by the similarity of genitures, he should unwittingly call, let us say, the Ethiopian white or straighthaired, and the German or Gaul blackskinned and woollyhaired, or the latter gentle in character, fond of discussion, or fond of contemplation, and the Greeks savage of soul and untutored of mind;  A fellow Dutchman thus should think twice before calling me wellmannered .
Not only the permanent issues like gender or descent play a role but also situations of the moment. Already some years ago I always had the impression that transits barely worked during several of my cycling or hiking holidays and lately I've been experimenting by ignoring transits and not thinking of astrology to see whether the transits still would work. I can't tell for sure yet what the outcome is.
Relating this to statistical research would be quite difficult though. 

Back to top 




You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum




.
Powered by phpBB © 2001, 2005 phpBB Group

