An analysis of birth and death dates for 161 persons who killed themselves in New York between 1969 and 1973 found no statistically significant common planetary transits.

To test the hypothesis that planetary astrological effects are weak but real, I conducted a small study of data collected by Ms Nona Press of dates of birth and death of suiciders. I acknowledge and thank Ms Press for kindly providing this data, and for referring me to her report on the study in her book New Insights into Astrology.

My interest in this material came from reading The Moment of Astrology by Geoffrey Cornelius, which explains that considerable analysis of natal data of this group in the 1970s failed to find any evidence for astrological predictions. My study sought to test the claim by astrologers that the planetary transit explanations in the book Planets in Transit by Robert Hand have an obvious real meaning for cycles in human psychology. My hypothesis was that if transits have an effect, then a reasonable size statistical analysis of extreme events, such as suicides, should show some patterns of common planetary timing. For example, the planets Mars and Saturn, traditionally known as malefics, could be thought to influence the mood of people considering suicide, such that more than normal numbers kill themselves at a particular point in the Mars-Saturn cycle, measured in relation to where planets were at the time the person was born.

I entered the 161 birth and death dates provided by Ms Press in hard copy into a spreadsheet, and then linked this to ephemeris data and calculated all transit angles for each planetary pairing. This produced a listing, each sorted by angle, of all individual planetary transits for the group at their date of death, as well as some planetary midpoint transits. Technical details are available on request.

The hypothesis posited that graphs of this group would look different from the null hypothesis of no effect. The transit graph for the Mars-Saturn midpoint at date of death is attached as a representative example. Looking at the position of this midpoint against the natal planetary positions of all 161 cases, the trend lines are very close to the population norm, shown by the diagonal straight line. All inner planet transits cluster around this line, while the outliers are artifacts caused by the short four year death date range (1969-73), during which the outer planets hardly moved. Any significant deviations from normal would be shown by lines appearing flatter for one part, showing bunching of suicides during a specific angle of the transit to the natal planetary positions. For example, if there were a tendency for people to kill themselves when the Mars-Saturn midpoint was opposite the point on the ecliptic where the Moon was when they were born, an outcome in rough conformity with Mr Hand’s predictions, the moon transit trend line would be horizontal at 180°, and it would be clearly visible away from the trend line. Instead it closely hugs the average population trend of the null hypothesis of no planetary effects. A discrepancy would show that more than average numbers killed themselves at one stage of the Mars-Saturn midpoint cycles against natal planetary positions. However all findings appear to be well within statistically probable limits, with roughly equal numbers at each angle and no detectable bunching. The same pattern emerged for other planetary transits.

This study shows that transit effects are too weak to be detected at this scale. However, it may be that the paths leading people to kill themselves are too diverse to be detected by this crude approach of comparing planet positions at dates of birth and death. Resolving data to exact moments could study the moon and also introduce further complex factors such as house positions. My interest was to see if a simple crude analysis would produce promising results, but it did not.

These findings do not disprove transit influence, but they show that any such influence is too weak to appear at this level of data. A larger epidemiological study of birth and death dates could increase the sensitivity of this same method. For example, looking at 100,000 heart attack victim dates of birth, onset and death might show significant results. If such data were available this method could quickly analyse to test for empirical validation of astrological predictions at population level.

Robert Tulip

Attached Thumbnails

 

I think you got the wrong forum. Try this one.

That forum is not empirical. The study described above is 100% empirical and scientific, producing a null finding. I am not interested in wild speculation without factual basis.

Are you arguing he's preaching to the choir?

__________________
‘To those who regard “crime fiction” as some sacred icon which must follow a rigid formula, I will always be the man who writes 18-syllable haiku.’Trying to make sense of computers, The Error Log.

Hi Robert,
I don't really have an interest in this sort of stuff, but you may find this interesting, as it seems to relate to the work you have done. Some fodder for falsification perhaps?

http://bourabai.narod.ru/landscheidt/consider.htm

This is a paper by Theodor Landscheidt about the golden section and it's manifestation in various aspects of life the universe and everything.

About halfway through, there's a section on the correlataions Gauquelin purportedly discovered between birth times and professions.

"Though Gauquelin's statistical work was state of the art and could be reproduced with new data, there was much criticism because the accumulations in the diurnal circle did not fall directly at or before the cardinal points rising, culmination and setting, but built up in between , and not even symmetrically. Yet I could show (Landscheidt, 1991) that the accumulations are exactly related to the cardinal points in the diurnal circle when the golden section is taken into account."

Thanks Stroller, I skimmed the article and it certainly shows the potential both for scientific research into these topics and the risks of veering off into speculation. Yes, phi is built into natural cycles and structures which display the Fibonacci sequence, and yes, it is possible to find planetary correlations with phi, eg those at http://solargeometry.com/Overview.htm . However, that is tangential to the point here, which is to describe a method for objective empirical testing of astrological claims. I found your avatar at http://www.orbitsimulator.com/BA/sbc5.GIF - the barycentre sans Sun sans Jupiter. Gravity simulator is a great site. You may be interested in the related discussion at http://www.bautforum.com/questions-a...e-gravity.html and Spiral Model of Solar System

I should say, I am not aiming here to 'preach to the choir' in Henrik's phrase. The baut pew sheet, at least in Phil Plait's expression, sees all astrology as intrinsically unfalsifiable and hence unscientific, whereas I am looking at falsifiable statistical analysis that holds out the faint prospect that weak regular correlations may be detectable when sufficiently large datasets are examined systematically.

Hi Robert,
Well spotted with the avatar. I am interested in matters barycentric, thanks for the links. The relative distances of the planets from the sun and relationships with the solar cycle is given a different and interesting treatment by the (unfortunately banned) Ray Tomes in these threads:
Harmonics Theory
Explaining Planetary Alignments Relationship to the Sunspot Cycle
He finds the outer planets on the nodes of a 160 minute lightspeed wave and the inner planets near a 5 minute wave.

Landscheidt has interesting things to say about jupiters motion, barycentric motion of the sun, and earth's geomagnetic index here:
http://bourabai.narod.ru/landscheidt/rotation.htm
he links this motion to El Nino here: (among many other of his papers)
http://www.john-daly.com/theodor/DecadalEnso.htm

All of which is as you say, tangential to your thread, although as a non-statistician it would seem that Landscheidt is reasonably rigorous in performing tests for confidence levels and null hypotheses.

__________________
Everything I need to know I learned through Googling.

Movement from General Science to Against The Mainstream is fine by me, although this thread was not started to present an ATM argument which disputes any mainstream science. I put this thread in General Science with the intention of respecting BAUT rule not to promote ATM in mainstream fora. This was in response to some earlier comments on an ATM thread to the effect that ATM should only be for promotion of propositions in conflict with existing scientific knowledge/consensus rather than for purely empirical discussion. I agree the policy clarification explained above by ToSeek is reasonable given the dubious content of the hypothesis and my underlying ATM intent. Now of course it comes under the thirty day closure rule with clock ticking to 7 Feb. May I reiterate how much I appreciate the opportunity afforded by the Bautforum for interaction on these difficult topics, which I generally find are subject to strong cultural repression, but which should be open to sceptical discussion. My agenda is to explore how rational empirical study can broaden the boundaries of knowledge regarding interaction between the earth and the solar system. This is a fine point, considering the weight of anecdotal superstition surrounding astrology, and the difficulty of cleaving a rational path within it.
My aim is to use mainstream statistical methods to analyse some major non-mainstream claims. This was the aim of Michel Gauquelin, who ended up killing himself in despair at the rejection of his findings about planetary correlations. It also may bear comparison to the work of Geoffrey Dean, who turned from advocacy of scientific astrology to the conviction that astrology is bunk, in the face of failure to find strong results. I do not share Dean’s pessimistic assessment, as I am hopeful that the problem is that planetary effects are simply much weaker than has hitherto been assumed in scientific studies.
As readers of my earlier BAUT ATM threads would know, I am sympathetic to astrology, but try to approach it from a scientific perspective. For convenient reference, here are links to these earlier threads, as well as to some comments I made on others’ threads and some non-ATM threads.
1. Science and Astrology
2. Cosmos and Psyche Planetary Theory
3. Planets and Rain
4. http://www.bautforum.com/questions-a...nus-earth.html
5. Planets and Earthquakes
6. Precessional Cosmology
7. Geocentrism
8. Jupiter influencing sunspots
9. Astronomical History
10. Explaining Planetary Alignments Relationship to the Sunspot Cycle
11. 178.867624
12. http://www.bautforum.com/general-sci...ros-cycle.html
13. http://www.bautforum.com/questions-a...e-gravity.html
14. http://www.bautforum.com/general-sci...-obsolete.html
15. Harmony of the Spheres
16. Spiral Model of Solar System
17. http://www.bautforum.com/questions-a...le-galaxy.html
18. http://www.bautforum.com/questions-a...2-vectors.html
19. 2009 Planet Calendar
20. http://www.bautforum.com/general-sci...l-pyramid.html

Robert Tulip

Yes, and posting "studies" of astrology gives it more credibility than it deserves.

I think it might be interesting to do a large scale comparison of a sub-population that is strongly agrarian and tied to seasonal variations in food supply vs. a highly urbanized group where food and other resources are available in steady supply year round. I would expect that if any patterns exist in the data, they would be stronger in the group where being born in a "lean" season has more impact to nutrition.


Regards,

Tes

Though really he's not, because he's made it clear himself that he does believe, or wants to believe, in astrology and similar things. I think he's willing to try to subject the idea to study, and see what happens. I think it's a laudable attitude, and a reason that this is probably a good place to post it.

__________________
As above, so below

Thank you Tesarra, your suggestion raises the problem I face as a private researcher, that of getting access to data. If a registry of births deaths and marriages gave me old date records for say 100,000 anonymous people, I would subject this data to the same tests described above. Many government agencies would have databases with suitable records, for example date of birth and date of accident for road deaths, etc. To my knowledge such testing has never been done before. It is very simple, but the ill-repute of astrology seems to pose barriers to doing it, and I have no idea how to obtain such data. Any suggestions?? Regarding your specific proposal, artifacts of the type you describe are common, given that terrestrial environmental cycles swamp planetary influences which are tiny if they exist at all. However, planetary transit data is not seasonal so I don't think the specific example you give would be a problem. My Planets and Earthquakes thread linked above contains a simple howler where I mistook an obvious artifact for a significant result. Robert

Just a quick google brought up Minnesota's online death certificate database:I dunno about other states, but in this MN case, it appears that the info is available--just go to the microfilm reader.

It's my view that a clear distinction can be made between ATM ideas which present potentially viable mechanisms for interaction and causation between planets and the earth or other solar system bodies, and ATM ideas which present correlative linkages of a statistical nature between the planets and human activity.

I would hope that the forum administrators are also able to see this distinction so that ideas falling within the former category are not subject to any blanket treatment of ideas falling in the latter category.

As an example, one idea might be that the planet with the magnetosphere bigger than the sun might have some effect on the geomagnetic index of the earth. Another might take this further and say that Jupiter position affects peoples predisposition to gamble or procreate or whatever. If the first idea has some solid statistical support and a potentially viable mechanism which could in theory be tested, it has falsifiable content which can at least potentially be dealt with via empirical observation. The second idea will probably only be amenable to statistical analysis, and the causative mechanism, although potentially explainable by the electro chemical biology of the brain, is not as accessile to empirical observation.

I would also hope that threads such as this one whose thrust is to treat scientifically data which may be subject to a-priori dismissal by those with a predisposition to skepticism in relation to the subject matter will be allowed. After all, they should be confident that a scientific appraisal of such data will always prove the thesis wrong, so what's the harm if it serves to reinforce their view and lends support to their contention?

If such a study were to find that some of the correlations being analysed have some validity, this should energise people on one side of the debate to seek viable causative mechanisms, and those on the other to prove that the statistical methods employed are in error.

Either way, learning is stimulated, lessons are learned, and progress is made, which is what it should be all about.

Thanks hhEb09'1, the USA is probably freeer with info than Australia. The couple of enquiries I have made have received no response. Perhaps I should pluck up the courage to write to Minnesota, or maybe to the Mormons, and explain clearly what I am doing. Lacking any institutional framework for a research ethics accreditation may be a problem, and I am not sure what university might help with that. The data is there, but they would need to provide it in the format I need, which is a spreadsheet with two columns listing dates of birth and death, and possibly a third column for cause of death, then as many rows as they can send. The prospect of a trip to Saint Paul in the winter to bunker down with a microfiche for three months to do data entry is somehow not that appealing.

Very good point, and precisely what I am trying to do. I have noticed that people in opposing camps tend to talk past each other, rather than engage in dialogue. Facts provide a path to a modus vivendi.

I am now analysing the suicide data to mine for bunching and identify possible groups of interest. Of course this presents the statistical problems of cherry-picking and post-hoc hypothesising, but it may reveal something worthwhile.

That's science for you.

I'm still waiting on the data that I need for my PhD. Twenty years ago, we thought it'd be any day then. I contacted some of the more influential researchers and thought I'd made some convincing arguments, but so far no go.