Research Studies: Tropical vs. Sidereal & Related

[Buz_Overbeck]

This new topic provides a space to present and discuss research and other studies pertaining to the two zodiacs. There are a lot of independent studies out there but with limited exposure. Hopefully, this forum will be a place where the researchers and their work can be made accessible.

One such researcher, who has done a tremendous amount of independent work, is Sari Metsovuori and I have invited her to participate in this forum. She promises to post some of her studies, results and methods and I'm hoping others will quickly follow.

Please feel free to comment and critique whatever research is presented here and submit work of your own.

Buz Overbeck
Moderator

[Sari_Metsovuori]

Thank you Buz Overbeck and Michael Erlewine for inviting me on this forum.

Yes, I’ve made some statistical studies where I’ve compared the zodiacs among other things. First I’ll describe my methods.

Methodology

I’ve taken 130 different groups out of AstroDatabank 4 sized from 66 charts to 1896 charts. There are professions, illnesses, personality features, etc. – whatever attributes ADB contains, and you can see the samples here http://koti.welho.com/jmetsovu/samples.htm . Then I’ve looked which of those groups have ‘a lot’ of something, be it a lot of planets in a certain sign or house, a lot of certain aspects, a lot of planets in certain dignities, etc. Generally I’ve used only the seven visible planets: Saturn, Jupiter, Mars, the Sun, Venus, Mercury and the Moon, the most obvious reason for that being that their sign positions are not generational, expect Saturn for some extent. When studying signs I could also have used the axis, but I’ve chosen not to do that in these studies. Instead of using a separate control group for every experimental group, I’ve used a single control group of 48 620 charts that’s made from artificial data but is based on real data of 4 862 random charts.

I’ve got some critique for studying several planets inseparably, and suggestions that studying the planets separately or only the ruler of the Ascendant or the ruler of the MC, etc. would give better results. I’ve studied separate planets and points too, but in these studies the main goal has been to find the common flavor of a sign, house, etc. No astrologer who sees a chart with a visible emphasis in a certain sign or house can ignore that in interpretation. Also using several planets gives us bigger sample sizes, because in that way when studying 100 charts we actually have a sample size of 7 * 100 = 700 planets, and the bigger the sample sizes, the more reliable the results. Of course the Sun, Mercury and Venus are never completely independent from each other, but I still consider this a useful option.

The first measure I’ve used is effect size which I’ve learnt from Geoffrey Dean’s article about Michel Gauquelin http://www.rudolfhsmit.nl/g-hist2.htm . The effect size measures how big a portion the difference between the experimental group and the control group is from the sample size, and it’s calculated by subtracting the expected value (got from the control group) from the observed value (got from the experimental group), and dividing the result with the difference between the sample size and the expected value. The formula is (observed value – expected value) / (sample size – expected value). The effect size is always between -1 and 1, and it can also be negative if the experimental group gives a smaller result than the control group. The effect size is especially useful in drawing diagrams.

The problem with the effect size is that it’s a lot easier to get big effects with smaller samples than with bigger ones, and so the results got from samples of different sizes are not comparable with each other. So we have to use another measure which I call “effect power” and which proportion the effect sizes to each other. The effect power is got by multiplying the effect size with itself and multiplying the result with the sample size. The formula is (effect size)2 * sample size. The effect power is always positive, so one has to artificially make negative the effect power values got from negative effect sizes for separating the negative results from the positive ones.

After that the effect size values got from different groups are simply put in numerical order, and then we can see which of the groups scored highest, which in turn scored lowest. The goal of the study is to find common themes from the highest scored groups which hopefully oppose those of the groups that scored lowest. If the hightest or lowest scored groups show no common themes whatsoever, then we can conclude that the technique in question doesn’t work. So, this kind of study is both quantitative and qualitative.

Sari Metsovuori

[Sari_Metsovuori]

Masculine and Feminine Signs

One of the simplest ways to define signs is to divide them into masculine and feminine ones. The masculine signs are expected to be active, extroverted, initiating, etc.; and the feminine signs are expected to be passive, reactive, introverted. We can study planets in signs using the tropical and sidereal Fagan / Bradley zodiacs, and look which of them shows better the masculine and feminine qualities of the signs. Then of course there is the question of what exactly defines masculinity and femininity, but that is another issue.

First the tropical zodiac. Here is the top-15 of the groups studied which had planets most frequently in the masculine signs:
1. PR people
2. Friendly / gregarious personalities
3. People with Western belief systems
4. Athletes
5. Basketball players
6. Magazine / newsletter writers
7. Top executives
8. Baseball players
9. Clerks / secretaries
10. People with metaphysical beliefs
11. Architects
12. Chemists
13. Attorneys
14. Lesbians
15. Producers

We can make some sub-groups to see the results more clearly. We have active people working in business and public relations:
- top executives
- PR people
- producers
- attorneys
- magazine & newsletter writers
- friendly & gregarious personalities.

We have also sportsmen:
- athletes
- basketball players
- baseball players.

How about the groups having a lot of planets in feminine signs? Here is the top-15 of those:
1. Composers / arrangers
2. Biologists
3. People moved a lot
4. Soldiers
5. Gay men
6. Long lived people
7. Religious / philosophical writers
8. Substance abusers
9. Textbook / non-fiction writers
10. Highly educated people
11. Humanists / philosophers
12. Expatriates
13. Spiritual leaders
14. People with vulnerable constitutions
15. Combatants

Let’s make some sub-groups again to get a better picture. At least we have scientists, philosophers and people who think about religious or spiritual issues:
- humanists & philosophers
- religious & philosophical writers
- spiritual leaders
- textbook & non-fiction writers
- biologists
- highly educated people
(-long lived people – I include this group here too)

Then we have people who may live a somewhat unstable life:
- substance abusers
- people with vulnerable constitutions
- people moved a lot
- expatriates
- soldiers
- combatants.

The last two groups are not especially feminine, but the general difference between the masculine and feminine groups is clear: the first one is active, extroverted, decisive, initiative, putting things into action. The latter group is reflective, more passive, thinking things through, more reactive than initiative. It would seem that the traditional masculine / feminine division shows well on the tropical zodiac.

What happens sidereally? Now the qualities are naturally reversed. Here is the top-15 list of groups that most often had planets in masculine sidereal signs:
1. Biologists
2. People with vulnerable constitutions
3. Spiritual leaders
4. Substance abusers
5. Poets
6. Nervous breakdown cases
7. Composers / arrangers
8. Humanists / philosophers
9. Humor writers
10. People with unusual diets
11. Private pilots
12. Expatriates
13. Rugby players
14. Lottery winners
15. Researchers.

Probably the most pronounced feature here is sensitivity. We have:
- people with vulnerable constitutions
- substance abusers
- nervous breakdown cases
- people with unusual diets (mostly vegetarians)
- humanists & philosophers
- spiritual leaders.

We have also artists:
- poets
- composers & arrangers
- humor writers.

Then we have a couple of groups of scientists:
- biologists
- researchers.

The 15 groups that had most often planets in feminine sidereal signs in this study were:
1. Friendly / gregarious personalities
2. Attorneys
3. PR people
4. People with exceptional minds
5. Farmers
6. Bankers
7. People with Western belief systems
8. Athletes
9. People with low IQs
10. Diplomats
11. Football players
12. Architects
13. Aggressive personalities
14. Race drivers
15. Biographers.

We have the same extroverted, social people here that we encountered with masculine tropical signs:
- friendly & gregarious personalities
- PR people
- bankers
- attorneys
- diplomats
- biographers.

We also have some sportsmen:
- athletes
- football players
- race drivers
- aggressive personalities.

The difference is clear, but the roles are curiously reversed, unless we consider sensitivity and introversion as archetypically masculine qualities. This study would seem to speak for the tropical zodiac, but in any case one thing is clear: we cannot apply tropical interpretations on the sidereal signs. If we want to use sidereal signs, we have to adjust the interpretations accordingly.

There are other areas where only one of the zodiacs can work, and one of them is dignities. Do people with several planets in their own signs or exaltations, or several planets in their detriments or falls, share similar features? And if they do, which zodiac shows those features more clearly? I will study that in the upcoming studies.

Sari Metsovuori

[Sari_Metsovuori]

Planets in their own signs

Traditionally when a planet is in its own sign, it’s considered strong and fortunate. It doesn’t need to rely on other planet’s resources and generosity but it’s independent and self-sufficient. It’s free to do whatever it wants. This should be visible in the life and maybe even personality of those who have many planets in their own domiciles. Let’s look how this manifests on different zodiacs.

First the tropical zodiac. Here is the top-15 list out of the 130 samples who most often had planets in their own signs:
1. Child prodigies
2. Private personalities
3. Nobel Prize winners
4. Opera singers
5. People moved a lot
6. Baseball players
7. Poets
8. TV personalities
9. Physicians
10. Publishers / editors
11. Thieves / financial criminals
12. Athletes
13. Farmers
14. Rapists
15. Diplomats

We have here:
- child prodigies, Nobel Prize winners, physicians, diplomats, opera singers
- TV personalities, publishers & editors, poets and
- baseball players, athletes.

All in all these are people from the higher levels of society who have succeeded with their own, often in-born giftedness. We have also a couple of groups who represent the self-sufficient and independent nature of the planets in their own signs:
- private personalities and
- farmers.

Who are people who lacked planets in their own signs in this study? Here’s the list:
1. Suicides
2. Humor writers
3. Lesbians
4. Astrologers
5. PR people
6. People served in the army
7. Real estate agents
8. Prisoners
9. Economists
10. Combatants
11. Artists / fashion designers
12. Executed people
13. Depressed people
14. Biographers
15. Healers / therapists.

Let’s make some sub-groups again. We have quite unfortunate samples:
- depressed people
- suicides
- prisoners and
- executed people.

The difference to the previous samples is clear here. Then we have
- people served in the army and
- combatants. We have also
- PR people
- economists and
- real estate agents. The former sub-group here has to fight for survival, and the latter sub-group has to fight for money.

When compared to the people with planets often in their own signs, here we have a sense of struggle and exertion. What came easily and naturally to the own-sign people, the not-own-sign people have to fight for. This is in harmony with the traditional interpretation for the planets in their own signs.

What happens with the sidereal (Fagan/Bradley) signs then? Do we have an even clearer manifestation of this kind of dignity? Here’s the list of people with planets often in their own sidereal signs:
1. Teachers
2. People in sex business
3. Biologists
4. Nervous breakdown cases
5. Nobel Prize winners
6. Race drivers
7. Rape / sex victims
8. Psychotic cases
9. Private personalities
10. People with high IQs
11. Combatants
12. Gorgeous looking people
13. Long lived people
14. Civil / political criminals
15. Rapists.

This really is a diverse group of samples. On the other hand we have:
- teachers
- biologists
- Nobel Prize winners
- people with high IQs
- gorgeous looking people and
- long lived people.

So far so good. But what about
- combatants
- civil & political criminals
- people in sex business
- rape & sex victims
- rapists
- nervous breakdown cases and
- psychotic cases then?

I cannot see a unifying factor. We can look at people with few planets in their own sidereal signs. Here’s the top-15 list:
1. Lesbians
2. Astrologers
3. Clerks / secretaries
4. Mathematicians / statisticians
5. Outdoor people
6. Physicians
7. Spiritual leaders
8. Artists / fashion designers
9. Physics
10. Top executives
11. Friendly / gregarious personalities
12. Columnists / journalists
13. Substance abusers
14. Diplomats
15. Government employees.

There’s even less common factors. We can try. We have
- mathematicians & statisticians
- physics and
- physicians. Then we have
- friendly & gregarious personalities
- diplomats
- government employees and
- top executives. And then
- spiritual leaders and
- astrologers. And finally
- columnists & journalists and
- clerks & secretaries.

This is beginning to look quite hopeless. Apparently here we have an example of a technique that doesn’t work. If one likes to find common themes in these groups with few planets in their own sidereal signs, which hopefully contradict those of the people with a lot of planets in their own sidereal signs, one has to stretch imagination to needless lengths.

So it seems that again the sidereal zodiac ended up with the short end of the stick.

Sari Metsovuori

[Buz_Overbeck]

Very interesting study, Sari. It seems like a theme is beginning to emerge...

Buz

[Kyle_Pierce]

Sari, it's interesting to compare the approach you have described in your recent posts with the one I've started writing about in my blog. My focus is on refining a hypothesis until it becomes as specific as possible, based on working with case studies. You are asking more general questions that depend on how natal chart features are related to an individual's vocation, personality traits, or attitudes, to name a few categories. You are casting a broad net and you need not feel insulted if someone tells you this is a "fishing expedition." That was my first reaction, but I haven't thought of a better way to deal with the questions you are asking, and now I see that your study is better thought out and more coherent in its approach than I did at first. Still, there is always room for improvement.

I would hope to see you try some alternative approaches to the control group and sampling issues, to see which of your findings show up or not, when trying one thing or another. For example, instead of using a fixed (though huge) control group for comparison with a wide range of experimental groups, you could of course generate controls from scratch for each experimental group.

There are likely to be some good reasons for having similar numbers of subjects in each pair of groups. The idea is to make the two groups in each pair as comparable as possible. But with your fixed control group of 40,000+ subjects, you are actually doing something a bit similar to what I've been doing. Even though you could call it a control group, it's more like you're comparing the experimental group to an unmatched group of charts, rather than controlling for anything -- that is, other than the irregular distributions of planets in the zodiac over certain periods of time.

In my dynamic synastry study, I'm generating a few thousand partner charts for each individual, for random dates within a specified range of years, and finding aspect contacts between that individual's solar arc planets and the generated partner charts. The contacts between the charts of the actual partners are then scored along with the generated charts, against a model that has been derived from working with a small collection of couples' charts. This is done for both partners.

The result of all this is an estimate of the probability that the pattern of aspect contacts between the original pair of charts could occur by chance. Not this precise pattern, though; it could include just the set of planets involved, or it could get more specific by including the types of aspects between planets. If you make this pattern specific enough, you get into the situation I described in my blog, where each case becomes statistically unique. This is not usually what we want to do, however.

It's good to have some discussions about these topics that have not been heard so much since the golden days of ACT. There are so many special areas of interest, but the question of how to gain better astrological knowledge -- which is the overriding question for researchers -- might deserve to be one of the bigger areas, because it could be such a useful resource for doing so many kinds of astrology.

[Sari_Metsovuori]

I would hope to see you try some alternative approaches to the control group and sampling issues, to see which of your findings show up or not, when trying one thing or another. For example, instead of using a fixed (though huge) control group for comparison with a wide range of experimental groups, you could of course generate controls from scratch for each experimental group.

There are likely to be some good reasons for having similar numbers of subjects in each pair of groups. The idea is to make the two groups in each pair as comparable as possible. But with your fixed control group of 40,000+ subjects, you are actually doing something a bit similar to what I've been doing. Even though you could call it a control group, it's more like you're comparing the experimental group to an unmatched group of charts, rather than controlling for anything -- that is, other than the irregular distributions of planets in the zodiac over certain periods of time.

Kyle, many thanks for your response. You’re right, ‘control group’ is a bit misleading word for a big bunch of unmatched charts. My goal in using that kind of big control data was mainly to mirror the normal distribution of planets. Most of my previous studies are done with specific control groups and they have advantages. Sometimes I felt though that these specific control groups mirrored the experimental data too closely. But the biggest advantage of using a fixed control data was that it fastened the study quite a lot!

I read in your blog about the synastry study you’re doing. The method of generating partners for every individual and looking the real contacts against those artificial ‘partners’ sounds credible indeed.

There are so many special areas of interest, but the question of how to gain better astrological knowledge -- which is the overriding question for researchers -- might deserve to be one of the bigger areas, because it could be such a useful resource for doing so many kinds of astrology.

This should be the main concern of the study: to gain better knowledge of how the patterns really work on charts. I’m not that interested in proving that astrology works (though that can be a byproduct, but some people will never believe it anyway) or even proving that a certain pet theory really works, but simply finding out what the planets, houses, signs, etc. really do at charts. It’s not always so evident as people think it is.

[Wayne_Turner]

Hello all,
This is my first post here, and this topic is naturally one that I looked into when I first began studying astrology back in 1978. I soon discoverd that both Tropical and Fagan/Bradley Sidereal positions were important, and in fact I also sometimes look at other reference points as well. I leave statistical studies to others, but my chart files and rectifications of historical charts were the keys that led to my use of both zodiacs. Robert Blaschke has posted one of the more powerful pieces of anecdotal evidence here regarding the conjunction of Obama's Tropical sun to McCain's Sidereal sun: viewtopic.php?f=22&t=7#p23 I have posted many messages over on Jim Eshelman's Solunars board, and if you're seriously interested in Western Sidereal I recommend that you browse through the topics there. Here is part of a post that I think qualifies as research, though a much larger sample and more comparisons would solidify the usefulness of this remarkable technique. If someone has seen many such contacts in the Tropical zodiac, it would suggest that both are important when using this technique. Originally posted here: http://solunars.net/viewtopic.php?f=28& ... nics#p1244

This is an excerpt from a little ebooklet I wrote, with other notes on why the 9th and 12th harmonics in the Fagan/Bradley Sidereal zodiac are powerful charts. Jim Eshelman added some powerful examples from his own files.
=====================================================================

A Time to be Born

When I was born the Sun was at sidereal longitude 202:17:28. When my father was born the Moon's position was 22:13:15, so that my Sun is closely opposing his Moon. The positions of my older brother's and step-mother's Suns are close together, since they were born on the same calendar day, January 11. The first Sun position is at 267:09:44, and the second is at 266:22:01, less than 1 degree away.

These connections seem intriguing enough, but there are other powerful contacts that become apparent only with some rather esoteric geometry and mathematics. In India there are derivative charts calculated from the natal chart using a method that divides the 360° ecliptic circle into smaller subsets. For example, by dividing the circle into 9 smaller "zodiacs" and folding them on top of one another, each mini-zodiac will span 40°, and every point that is a multiple of 40 degrees apart will be in conjunction in the new subsidiary chart. The same will be true of other divisions, such as 12 or 16. In the West these derivative charts are called harmonic charts. To calculate such a chart you can multiply the longitude of the original point by the circle division, then use the new longitude in your harmonic chart, making sure that it is less than 360° by eliminating any excess from the product. Thus a point at 10° in the root or radix chart will be at 90° in the ninth harmonic chart (10x9) and at 120° in the twelfth harmonic chart.

By multiplying the sidereal longitude of my father's natal Moon by 12 it is found to have a harmonic position of about 266:39, only 17 arc minutes from his second wife's natal Sun. The twelfth harmonic position of his Moon is also about 31' from the Sun's position in his first child's horoscope--my older brother. My 12th harmonic Sun is at 267:29:39, about 20' from my brother's Sun, and just beyond the 1° orb that I allow with this technique for a conjunction to my step-mother's Sun. I chose to live with these three people when I was a teenager instead of with my mother and step-father.

Here was another powerful, personal confirmation that made it even more difficult for me to believe that astrology was mere fantasy. In later years I found many other confirmations of the power of this technique, which is actually not used by most astrologers.

It is also worth pointing out that the exact absolute longitude of the radix position has a critical bearing on the accuracy of the method. A difference of only 5' in the initial position becomes 45' in the 9th harmonic and 60'—a full degree--in the 12th. Since an orb, or zone of influence, of only one degree is allowed, the initial position must be based on a sidereal longitude that gives a best fit when comparing charts. Western Sidereal positions--discovered in the 1950's--work very well, while those used most extensively in India do not. Tropical positions also do not appear to be effective.

===================================================================================

Here is more data from my old files, with all positions in the Fagan/Bradley Sidereal zodiac, conjunctions with orb of less than one degree:

An astrologer friend had natal Uranus at 26GE21, with a ninth-navamsha harmonic position at 27TA06, conjunct her natal Mercury at 27TA33.
My natal Moon at 25LE49 gives a 12th H of 9AQ45, conjunct the above friend's natal Moon at 9AQ06.

A girl I was quite fond of had natal Venus at 12CP09, 12th H 25LE52, conjunct my natal Moon at 25LE49.

Another female friend had natal Moon at 2GE51, 9th H 25LI40, conjunct my ASC at 25LI35.
Close female friend (and friend of the above girl) natal Venus 17LE08, 12th H 25LI32, also conjunct my natal ASC at 25LI35.

Niece (by marriage) natal Moon 22SA30, 9th H 22LI26 conjunct my natal Sun 22LI17.

Co-worker who became a cosmetologist, natal Moon at 7VI23, 9th H 6PI27, conjunct her natal Venus at 6PI59.

Numerologist friend had natal Mercury at 16LI22, 12th H 16LI24 conjunct her natal Mercury!

Choreographer and dancer had her natal Neptune at 18LI33, 9th H 16PI56 conjunct her natal Mercury at 16PI57!

Former sister-in-law, natal Moon at 12CP07, 12th H at 25LE21 conjunct my natal Moon.

Woman's natal Moon at 16CA33, 12th H 18LI31 conjunct her daughter's natal Sun at 18LI43.

Using 15 data the combined error totaled minus 121 arcseconds, with an average error of about minus 8", meaning that adding 8" to the commonly calculated Fagan/Bradley ayanamsha would give a best fit to this small sample. I consider this to be a neutral result, and do not think that changing the ayanamsha would be warranted without much further work. Donald Bradley came up with his value back in 1957 I believe by working with more than one hundred ingress charts, mostly Lunar and Solar sidereal Capricorn ingresses.

I would be very interested in what others find using this approach.
Regards,
Wayne

[Buz_Overbeck]

Interesting post, Wayne. What's even more interesting is the number of female friends you have! It seems like the 12th harmonic is the standout in your work. Have you refined a "meaning" for the 12th and, if so, what is it?

Also, you say: Western Sidereal positions--discovered in the 1950's--work very well, while those used most extensively in India do not.. I assume you're meaning Indian Sidereal positions based on one of their many ayanamsas. Which ayanamsa(s) did you look at?

In any case, you make a strong anecdotal argument for Fagan/Bradley. Thanks for posting it.

[Ray_Murphy]

Sari wrote:
I’ve taken 130 different groups out of AstroDatabank 4 sized from 66 charts to 1896 charts. There are professions, illnesses, personality features, etc. – whatever attributes ADB contains, and you can see the samples here http://koti.welho.com/jmetsovu/samples.htm [non-working link] Then I’ve looked which of those groups have ‘a lot’ of something, be it a lot of planets in a certain sign or house, a lot of certain aspects, a lot of planets in certain dignities, etc.

I've been doing the same sort of thing with a semi-automated system for Occupations, Hobbies, Interests, Traits and Events, then getting the program to make a printout of the scores in percentage order and also highlight those categories that have the highest Chi squared scores. Any category that doesn't get a nominally interesting X^2 score is ignored and so are categories that get super-high scores - they are always categories with a very low number of samples, but their inclusion seems valuable, especially when the category rates highly percentage wise - along with another similar category with a lot more data.

An example might be a small number of "Military pilots" rating highly - along with a large number of "Commercial Pilots" for a certain astro-factor. Normally it is quite pointless to do statistical work with small numbers of samples because research so far has shown us (or shown me at least) that nothing in astrology works a lot of the time if we are looking at a single factor and a single trait or event. It appears to me that the most we'll ever see for that sort of thing is <10% increase or decrease - so if we find our research is giving us "50% above the expected score" we can be fairly certain that the finding is not representative of the real situation.

Don't get me wrong here - I'm not against using stats to try and get into the ballpark. I do it all the time.