Astrology and Statistics

[Kort]

Here's an interesting article on the use of statistics in science from ScienceNews.org:

'Odds Are, It's Wrong: Science fails to face the shortcomings of statistics'

http://www.sciencenews.org/view/feature/id/57091/title/Odds_Are,_Its_Wrong


“During the past century, though, a mutant form of math has deflected science’s heart from the modes of calculation that had long served so faithfully. Science was seduced by statistics, the math rooted in the same principles that guarantee profits for Las Vegas casinos. Supposedly, the proper use of statistics makes relying on scientific results a safe bet. But in practice, widespread misuse of statistical methods makes science more like a crapshoot.

It’s science’s dirtiest secret: The “scientific method” of testing hypotheses by statistical analysis stands on a flimsy foundation."


Is it useful or even appropriateto drag astrology into this mess?  My guess is that astrology is from a different realm that can't be measured and categorized by statistics.

[Ray Murphy]

Here's an interesting article on the use of statistics in science from ScienceNews.org:

'Odds Are, It's Wrong: Science fails to face the shortcomings of statistics'

http://www.sciencenews.org/view/feature/id/57091/title/Odds_Are,_Its_Wrong

[.....]

Is it useful or even appropriateto drag astrology into this mess?  My guess is that astrology is from a different realm that can't be measured and categorized by statistics.

Astrology as a whole cannot ever be tested, only individual claims could be, but there's
one big problem - we don't have specific claims TO test yet.

[Kort]

Astrology as a whole cannot ever be tested, only individual claims could be, but there's
one big problem - we don't have specific claims TO test yet.

In natal astrology time well spent would mean time of careful attentiveness to each individual case – seeing what that chart has to say about that person.  Lumping people together in groups is always of limited – and rather dimwitted – use, in any case.  Which brings us to simplistic identification with a sun sign  . . . . . .

[Ray Murphy]

Astrology as a whole cannot ever be tested, only individual claims could be, but there's
one big problem - we don't have specific claims TO test yet.

In natal astrology time well spent would mean time of careful attentiveness to each individual case – seeing what that chart has to say about that person.  Lumping people together in groups is always of limited – and rather dimwitted – use, in any case.  Which brings us to simplistic identification with a sun sign   . . . . . .

It doesn't matter what we do, as long as we can demonstrate that we can do what
we SAY we can do.

Do we ever see anyone saying what they can do? Yes of course we do and skeptics
of astrology believe us. They have no doubt at all that we can convince people
we've given good readings, however that's not the real issue. They (and lots of
other astrologers) would like us to prove we are actually doing it.

[Ray Murphy]

[...] Lumping people together in groups is always of limited – and rather dimwitted – use, in any case.  Which brings us to simplistic identification with a sun sign  . . . . . .

If astrologers won't tell each other what works best either in literature or at conferences,
and demonstrate it, then maybe the researchers can do it.

[Lopaka]

Actually, I couldn't disagree more with the premise that astrology cannot be tested by statistical methods. The fundamental problem is that most people, and an unfortunately large number of "scientists", do not understand statistics, its methods, application and limitations. That, however, is a problem,  with the audience, not with the techniques or the sciences that act in reliance thereon.

One of the most misunderstood limitations is that statistics applies, not to individuals, but to groups. Moreover, it is limited by the validity of the sampling universe used by the researcher. Astrology will be validated, if at all, only when we can demonstrate that it has mass application. If we can subsequently refine it to apply to individuals, then well and good.

One of the strongest areas where objective verification of astrological prediction may be found is in the arena of financial astrology. The activities of financial markets is one of large numbers of individuals acting in a common arena. People buy and sell tens of millions of shares of stocks each day. The number of businesses whose shares are traded number in the thousands. These sampling populations lend themselves eminently well to statistical analysis, and financial astrology may make predictions that are confirmed or refuted in very short order. The predictions of the better practitioners tend to be veified more often than not.

Medical astrology also lends itself to this technique. Unfortunately, privacy concerns and our tainted reputations often limit the research astrologer's access to this information.

I point to the Gauquelans' work as a hallmark of statistical astrological work, work that the practitioner ignores at his or her peril.

Finally, statistics is an established mathematical discipline. Its techniques are well established, and have been since the early eighteenth century when Gauss began to publish on the discipline. Research premised on statistics may be assessed, critiqued, defended and either confirmed or not, as the case may be, based upon well established and objective criteria. This is a big plus to any area of study that wants to confirm its legitimacy. 

When all is said and done we must admit that serious research astrologers should master statistics, not avoid it If we chose to ignore it and other strict scientific methodology we merely justify the criticism of those who chose to dismiss us out of hand. We prove that we study a mere pseudo science. We remain a joke, an intellectual irrelevancy. Our study is capable of and deserves so much more.

Aloha

[Ray Murphy]

Actually, I couldn't disagree more with the premise that astrology cannot be tested by statistical methods. The fundamental problem is that most people, and an unfortunately large number of "scientists", do not understand statistics, its methods, application and limitations. That, however, is a problem,  with the audience, not with the techniques or the sciences that act in reliance thereon.

One of the most misunderstood limitations is that statistics applies, not to individuals, but to groups. Moreover, it is limited by the validity of the sampling universe used by the researcher. Astrology will be validated, if at all, only when we can demonstrate that it has mass application. If we can subsequently refine it to apply to individuals, then well and good.

The sampling universe for astrological research for tropical astrologers, is mostly North America
and a part of Europe. We don't have any birth and event data for most of the world and only
microscopic amounts for some large regions.

One of the strongest areas where objective verification of astrological prediction may be found is in the arena of financial astrology. The activities of financial markets is one of large numbers of individuals acting in a common arena. People buy and sell tens of millions of shares of stocks each day. The number of businesses whose shares are traded number in the thousands. These sampling populations lend themselves eminently well to statistical analysis, [....].

It' still mainly parts of Eurpope and North America for most of the activity.
I heard on radio a few decades ago that 90% of all the money in the world was owned by 10%
of middle aged men in North America - so presumably it is still they and a small number of
newcomers who are shifting most of the stockmarket shares around.

[....]

I point to the Gauquelans' work as a hallmark of statistical astrological work, work that the practitioner ignores at his or her peril.

I'm still not convinced about the Gauquelin's work because the 'Mars Effect' for sportsmen hinged
on only 40 extra cases, and if the pairing of Sectors hadn't been done, then none of the observations
would have been statistically significant to an impressive degree (or at all).

Also the fact that replication was done with some of the work is not a surprise to me any more,
because I found a few years ago that trends routinely last quite a few years then disappear.

At least it's clear that the Gauquelins didn't cherry pick the Sportsmens data (by year) because
the Mars Effect didn't exist in the dataset for the first several hundred samples. If they had omitted
the earlies data, the results would have been better.   [....]

When all is said and done we must admit that serious research astrologers should master statistics, not avoid it If we chose to ignore it and other strict scientific methodology we merely justify the criticism of those who chose to dismiss us out of hand. We prove that we study a mere pseudo science. We remain a joke, an intellectual irrelevancy. Our study is capable of and deserves so much more.

Aloha

We can bypass statistics altogether and demonstrate anything that we think is solid. That's the
only way we ever get proof about anythng.

[Lopaka]

Ray:

Your observations about our limited sampling universe, while apt, merely illustrates an experimental problem we face and does not speak to the underlying statistical methodology. It may, in fact, be one of the roadblocks we face in gaining wider acceptance and merely begs for more work, not abandonment of a valuable analytical technique.

Re the Gauquelins, their work went far beyond sports figures and Mars. It may not be complete, but stands as a valuable contribution to the advancement of astrology.

You say, "We can bypass statistics altogether and demonstrate anything that we think is solid. That's the
only way we ever get proof about anythng." Well, first, demonstation and proof are different matters. A post hoc arguement may be an apt demonstration of a proposition; but it proves nothing. Second, you seem to say that abandoning statistics is the only way to prove anything? But maybe I misunderstand; I honestly don't understand what you're saying here. Finally, your statement approaches saying that anything we think may be proven to be true. If so, astrology is relegated to the realm of carnival fortune telling. If so perhaps it should be outlawed.  You may remember the lyrics from the 5th Dimension's song, "Age of Aquarius":

When the Moon is in the Seventh House
And  Jupiter aligns with Mars
Then Peace will guide the planets
And Love will rule the stars!

Huh?? What an insipid,  meaningless bowl of gruel. The Moon is in the Seventh House for about 2 hours every day, and I don't see alot more peace and love now than before. Then, Jupiter "aligns" with Mars". What's that all about? Aligns? And Mars as a planet of peace and love? But this nonsense brought many people to the study of astrology; and abandoning science could allow us to "prove" it. A brief read through the "Astrology" advertisements of the Farmers' Almanac will illustrate my point. "Madam Suzie will bring love, riches and banish bad luck." This is only one of about 50 similar advertisements; but astrology? Well, .......if science and intellectual rigor are abandoned it could be.

Science is hard; and it should be. But it will be professional suicide for astrology to abandon it.

[Ray Murphy]

Ray:

Your observations about our limited sampling universe, while apt, merely illustrates an experimental problem we face and does not speak to the underlying statistical methodology. It may, in fact, be one of the roadblocks we face in gaining wider acceptance and merely begs for more work, not abandonment of a valuable analytical technique.

I'm not suggesting abandoning statistics. It's just that ultimately we need to demonstrate that things held up reasonably consistently over a long period of time. Like, it's no good having a good p-score if it only applies to the second half of the data.

Re the Gauquelins, their work went far beyond sports figures and Mars. It may not be complete, but stands as a valuable contribution to the advancement of astrology.

We need to see if the observations hold up over a long period when the data is placed in birth (or event) order. I found quite strong patterns in the Gauquelin data that fizzled out, so I'm half expecting it to happen with some of the published findings too when more data is added to it in birth sequence.


Pubic domain image

You say, "We can bypass statistics altogether and demonstrate anything that we think is solid. That's the
only way we ever get proof about anythng." Well, first, demonstation and proof are different matters. A post hoc arguement may be an apt demonstration of a proposition; but it proves nothing. Second, you seem to say that abandoning statistics is the only way to prove anything? But maybe I misunderstand; I honestly don't understand what you're saying here. Finally, your statement approaches saying that anything we think may be proven to be true. If so, astrology is relegated to the realm of carnival fortune telling. If so perhaps it should be outlawed.  

Well take my observation about people gravitating towards tropical astrology more often than expected for centuries if they have Sun in tropical Aquarius. That observation doesn't need stats - it's obvious when anyone looks at the graph and the scores. That sort of continuity is what we need for all good observations. Having a good p-score is not good enough. Sure, anyone can apply any stats test they like, or argue with each other about p-scores but straight percentages show what really happened.

[Lopaka]

Ray:

On the one hand we agree. A "good p-score" is not sufficient, in and of itself. What is the standard deviation? How was the sampling conducted? What other factors may affect the outcome? What are the kurtosis and skew of the distribution curves? Has the curve a fat or skinny tail, or is it truly Gaussian? On the other hand, we don't. Your observation about folk with Aquarian Suns is an excellent example of my point about post hoc analyses. By itself your observation is merely anecdotal. it proves nothing, although it suggests much. I would want to see what percentage of people with Aquarian Suns gravitate toward tropical astrology, not merely an observation that a large number of them do. How many remain with siderial astrology? What percentage, based upon a reliable population sample, prefer each? How much duplication exists? How are these preferences expressed and do they differ from other Sun Sign natives? Is the difference statistically significant? Your observations suggest an area for fruitful research, but they do not prove a hypothesis. Neither do they refute the applicability of statistics, and especially with respect to an area with as many variables as astrology.

Aloha.

[Ray Murphy]

Ray:

On the one hand we agree. A "good p-score" is not sufficient, in and of itself. What is the standard deviation? How was the sampling conducted? What other factors may affect the outcome? What are the kurtosis and skew of the distribution curves? Has the curve a fat or skinny tail, or is it truly Gaussian? On the other hand, we don't. Your observation about folk with Aquarian Suns is an excellent example of my point about post hoc analyses. By itself your observation is merely anecdotal. it proves nothing, although it suggests much.

It proves the claim that people with Sun in tropical Aquarius gravitated towards tropical astrology and
had their names listed in Astrodatabank consistently more often than the non-Aquarian average.
Obviously it doesn't prove that there is a correlation between the movement of celestial bodies and
human affairs, but that was never claimed anyway.

What is the standard deviation? - Who cares when we've got something solid like this to follow up on.
How was the sampling conducted? - Straight out of Astrodatabank version 3 (non-updated online).
What other factors may affect the outcome? - Possible bias towards Aquarius by data collectors.
What are the kurtosis and skew of the distribution curves? - Who cares if it keeps getting better
- as it has for the last 100 years without anyone reporting it until I did.

I would want to see what percentage of people with Aquarian Suns gravitate toward tropical astrology,
not merely an observation that a large number of them do. How many remain with siderial astrology?

Good - that's why I do these surveys so people will be motivated to start looking and trying to demolish
many observations like this.

What percentage, based upon a reliable population sample, prefer each? How much duplication exists? How are these preferences expressed and do they differ from other Sun Sign natives? Is the difference statistically significant? Your observations suggest an area for fruitful research, but they do not prove a hypothesis. Neither do they refute the applicability of statistics, and especially with respect to an area with as many variables as astrology.

Aloha.

Well like I said, I wasn't trying to prove an hypothesis. I 've simply provided the facts and can prove
them, or they can be checked by virtually anyone who has an interest because it is all explained in
very simple language.

To save time later on - this is not about Sun in tropical Aquarius and 11 other Sun signs - it's about
Aquarians and non-Aquarians.

If anyone knows of anything better than this in astrology for 1/12 of the zodiac - yell out.


Public domain image

[Ray Murphy]

Actually, I couldn't disagree more with the premise that astrology cannot be tested by statistical methods. The fundamental problem is that most people, and an unfortunately large number of "scientists", do not understand statistics, its methods, application and limitations. That, however, is a problem,  with the audience, not with the techniques or the sciences that act in reliance thereon.

One of the most misunderstood limitations is that statistics applies, not to individuals, but to groups. Moreover, it is limited by the validity of the sampling universe used by the researcher. Astrology will be validated, if at all, only when we can demonstrate that it has mass application. If we can subsequently refine it to apply to individuals, then well and good.

One of the strongest areas where objective verification of astrological prediction may be found is in the arena of financial astrology. The activities of financial markets is one of large numbers of individuals acting in a common arena. People buy and sell tens of millions of shares of stocks each day. The number of businesses whose shares are traded number in the thousands. These sampling populations lend themselves eminently well to statistical analysis, and financial astrology may make predictions that are confirmed or refuted in very short order. The predictions of the better practitioners tend to be veified more often than not.
[...]

If there is a correlation between overhead transits and the various stockmarket movements,
it should be be visible to some extent in the data of each. This would mean finding patterns for one
stockmarket - and then seeing how many others moved in the same ways. It can also be done
with individual stocks. I wrote a program to do this a few years ago but haven't shovelled much
data into it yet.

One thing I noticed when I was using that program, is that trends which are related to transits
come and stay a while and then go, so it is quite a good idea to follow those trends while they
are working. I think this is what "artificial intelligence" programs that are based on astrology are
detecting

[Lopaka]

Dear Ray:

This is some very interesting data, and I would like to follow up on it with you. Moreover, and despite your protests, it appears that you are making a good start at a statistical analysis in support of your hypothesis. You may not have completed it, but it looks like a promising beginning.

[Ray Murphy]

Dear Ray:

This is some very interesting data, and I would like to follow up on it with you. Moreover, and despite your protests, it appears that you are making a good start at a statistical analysis in support of your hypothesis. You may not have completed it, but it looks like a promising beginning.

As you can see from the Gauquelin data example above, it may well be pointless to
bother with stats until we're well past 1000 datapoints if there are only 6 bars in a
graph, and 2000 datapoints if there are 12 for things like occupations - and if the
astrologers with Sun in tropical Aquarius observation doesn't hold up, it will be a lot
more than 2000. Currently people involved in stats are still saying we need a minimum
of 60 for chi squared if we are working with 12 factors like sun signs, but it's far, far
away from that for nearly everything in astrology.

I think, going by what I've seen in hundreds of projects like this, that it will be a
different story for very specific traits and events and perhaps hobbies - where luck
has a lot less to do with the ability to do those things.

Disappearing trends are quite common.

[Kort]

The large number of astrologers with the Sun in Aquarius may be meaningful, but differently than is commonly thought.  Around half of those Sun in Aquarius astrologers will also have Mercury in Aquarius.  Mercury is the traditional planet of astrology and astrologers.  Uranus came along and was given astrology and was also given Aquarius to 'rule', pushing out Saturn.  Since Uranus now is said to rule both Aquarius and astrology, Aquarius itself is now often associated with astrology.  A new planet and a change to the rulership scheme has changed Aquarius.   

Aquarius is a fixed air sign.  It is (still) ruled by Saturn.  Aquarius likes structure and organized thought, and is oriented toward humanity and aiding the collective.  Despite all the New Age dingbat ways these days, astrology is still based on a logical structure.  A chart needs to be approached logically and systematically for the insight to come through.  Conservative boundary-defining Saturn and fixed Aquarius bring structure to the realm of the mind. 

The last signs, Aquarius and Pisces, do take us further out from the limited scope begun with Aries, but whereas Pisces is a releasing to a more unlimited realm, Aquarius could be an opening to that realm through the application of structured thought – astrology.  However, it looks rather significant that around 89% of astrologers do not have Sun in Aquarius.  In other words, this number thing is pretty much a waste of time.