What do you mean, “test” astrology?

That was the actual question someone asked me recently. I had said that whenever astrology is tested, it fails the test. Her reply was, what do you mean, “test” astrology? I said, well, compare the predictions made by astrology, to what actually transpires. (i.e., see if it works.) It seemed a pretty obvious reply. It seemed to confuse her, though.

Astrologers don’t like testing astrology to see if it works. Can’t say I blame them, because it doesn’t work. Think of this. An expert astrologer draws up your detailed chart based on your accurate natal data. You are given this horoscope, and two others of different people. Would you be able to tell which one is yours? Because if not, what does it mean to say that astrology works? Whenever this test is done, people are not able to do this with any greater probability than pure chance. (i.e. one in three would get it right.)

Possibly the most detailed test of astrology using this type of method, was performed by Shawn Carlson. His paper, “A Double-blind Test of Astrology”, was published in the peer reviewed scientific journal Nature, in 1985. The interesting thing is that the San Francisco chapter of the National Council for Geocosmic Research recommended the 28 professional astrologers who took part, and (with Carlson), designed the tests. They also predicted, in advance, what they would consider to be a successful test.

Two tests were performed:

Test #1: Astrological charts were prepared for 83 subjects, based on natal data (date, time and place of birth), provided by the subjects. Each subject was given three charts: one chart based on their own natal data, and two charts derived from natal data of other people. Each subject was asked to identify the chart that most correctly described them. In only 28 of the 83 cases, the subject chose their own chart.  This is the exact success rate expected for random chance. The astrologers predicted that the subjects would select their own chart more that 50% of the time.

Test #2: 116 subjects completed California Personality Index surveys and provided natal data (date, time and place of birth). One set of natal data and the results of three personality surveys (one of which was for the same person as the natal data) were given to an astrologer who was to interpret the natal data and determine which of the three CPI results belonged to the same subject as the natal data. In only 40 of the 116 cases, the astrologers chose the correct CPI. As with test #1, this is the exact success rate expected for random chance. The astrologers predicted that they would select the correct CPI profiles in more that 50 per cent of the trials.

Conclusion by Carlson:

"We are now in a position to argue a surprisingly strong case against natal astrology as practiced by reputable astrologers. Great pains were taken to insure that the experiment was unbiased and to make sure that astrology was given every reasonable chance to succeed. It failed. Despite the fact that we worked with some of the best astrologers in the country, recommended by the advising astrologers for their expertise in astrology and in their ability to use the CPI, despite the fact that every reasonable suggestion made by advising astrologers was worked into the experiment, despite the fact that the astrologers approved the design and predicted 50% as the "minimum" effect they would expect to see, astrology failed to perform at a level better than chance.

"I have not yet received a serious scientific challenge to the paper. The newsletter of the American Federation of Astrologers Network published a response in January (1986). I was very disappointed to see that it largely consists of personal attacks. Its few substantive criticisms are attributable to ignorance of the experiment, of the CPI, and of basic scientific methodology."

(My bold.)

So the astrologers failed their own test. Does this mean they gave up astrology as being useless? Of course not: they are totally closed minded to the possibility that astrology doesn’t work.

Of course, in science one test alone doesn’t decide things. That’s why I have summarized below an additional 36 similar types of tests that astrology failed.

36 Other Tests Astrology Failed

Gauquelin, M.
Zodiac and Personality: An Empirical Study
Skeptical Inquirer, 6:3, 57
1982

Compiled personality profiles from biographies of 2000 sports figures, actors, scientists, and writers.  Compared these profiles with personality traits associated with the sign of the sun, moon, and ascendant according to eight astrology texts.  No correlation was found using either the sidereal or tropical zodiac.

Press, N., Michelsen, N.F., Russel, L., Shannon, J., Stark, M.
The New York Suicide Study
Journal of Geocosmic Research, 2, 23-47
1978

Examined records of suicides in NYC from 1969 to 1973. Selected all suicides who were born in NYC and for which birth data was available. This resulted in 311 suicide cases. 

For each of these, a control subject was randomly chosen who was born in the same borough and year.  The suicides and matching controls were divided into three groups according to year of suicide.

A computer program was used to test 100,000 different astrological factors in each of the 622 birth charts for significance between suicide and control groups.  None of the factors consistently correlated with the suicide cases.

Culver, R.
Sun Sign Sunset
Pachert
1979

Van Deusen, E.
Astrogenetics
Doubleday
1976

Culver, R., Ianna, P.
Astronomy Quarterly, 1, 147
1977

The above three references examined the correlation between sun sign and over 60 occupations.  The results of all three were negative -- no correlation was found between occupation and sun sign.

Dean G., Mather, A.
Recent Advances in Natal Astrology
p113
The Astrological Association
1977

Silverman, B., Witmer, M.
Astrological Indicators of Personality
Journal of Psychology, 87, 89
1974

Per Dalen,
Season of Birth
American Elsevier Publishing
1975

Pellegrini, R.,
The Astrological Theory of Personality
Journal of Psychology, 85, 21
1973

The above 4 references all found no correlation between sun sign and personality traits as measured by standardized psychological tests, mostly the California Personality Inventory (CPI). However, Pellegrini found a slight correlation between the CPI femininity index and season of birth.

Illingworth, D., Syme, G.
Birthday and Femininity
Journal of Social Psychology, 103, 153
1977

Tyson, G.
Astrology or Season of Birth: A 'Split-Sphere' Test
Journal of Psychology, 95, 285
1977

These two studies found no correlation between sun sign and personality traits measured by the CPI, including the femininity index.

Mayes, B.,  Klugh, H.
Birthdate Psychology: A Look at Some New Data
Journal of Psychology 99, 27
1978

Compiled natal charts and results of Minnesota Multiphasic Personality Inventory and the Leary Interpersonal Check List for 196 subjects. Compared 13 personality traits with sun signs, signs and houses of the moon and 8 planets, and with five planetary aspects.  No correlations were found.

Mayo, J., White, O., Eysenck, H.
An Empirical Study of the Relation between Astrology Factors and
Personality
Journal of Clinical Psychology, 105, 229
1979

Jackson M.
Extroversion, Neuroticism, and Date of Birth: A Southern Hemisphere
Study
Journal of Psychology, 101, 197
1979

These two studies found correlations between astrological factors and the Introversion/Extroversion index of the Eysenck Personality Inventory.

Veno, A., Pammunt, P.
Astrological Factors and Personality: a Southern Hemisphere
Replication
Journal of Psychology, 101, 73
1979

Failed to duplicate the correlation found above.

Pawlik, K.,  Buse, L.,
Self-attribution as a Differential Psychological Moderating Variable
Zeitschrift fur Sozilpsychologie, 10, 54
1979

Showed that the correlation above could be explained by the fact that some of the subjects knew what the expected results would be for their astrological signs.

Eysenck, H.,
Astrology: Science or Superstition?
Encounter, Dec 1979, p85

Jackson, M., Fiebert, M. S.
Introversion-Extroversion and Astrology
Journal of Psychology, 105, 155
1980

Saklofske, D., Kelly,McKerracher, D.
An Empirical Study of Personality and Astrological Factors
Journal of Psychology, 110, 275
1982

These three studies found no correlation between astrological factors (sun and planetary) and personality, including the introversion/extroversion index of the Eysenck Personality Inventory.

Culver, R., Ianna, P.
Astrology: True or False, p215
Prometheus
1988

A double blind test of astrologer John McCall was organized at the University of Virginia by Charles Tolvert and Philip Ianna.  McCall claimed an 80 percent success rate in choosing the correct natal horoscope for a subject from three false ones.  Twenty-eight subjects were chosen according to McCalls requirements (naturally born caucasians).  McCall had 7 successes out of 28 trials, exactly the number predicted by chance.

Silverman, Bernie I.,
Contemporary Astronomy by J. Pasachoff, cf p437
W. B. Saunders

Kop, P., Heuts, B.
Journal of Interdisciplenary Cycle Research 5, 19
1974

The above 2 studies found no correlation between marriage/divorce rate and sun sign combinations in the state of Michigan and the city of Amsterdam respectively. John McGervey Physicist Western Reserve University Found that the sun signs of 6,000 politicians and 10,000 scientists were randomly distributed.

Dean, Geoffrey
(don’t have reference)

Astrological readings were done for a groups of subjects.  The content of some of the readings were reversed (changed phrases describing the subject to their opposites).

Subjects reported that both the reversed and normal readings applied 95 per cent of the time

Gauquelin, M.
L'Influence des Astres, Etude Critique et Experimentale
Dauphin Press
1955

Found no correlation between occupation and the zodiac signs containing Mercury, Venus, Mars, Jupiter, Saturn, and the Moon at the time of birth.

Gauquelin, M.
The Cosmic Clocks, p84
Henry Regnery Co.
1967

Found random distribution of the house containing Saturn for successful individuals, and the house containing Mars for murderers.

Barth, J., Bennet, J.
Leonardo 7, 235
1974

Found no correlation between occupation, medical problems, height, longevity, and the zodiac signs containing Mercury, Venus, Mars, and Jupiter at the time of birth.

Culver, R., Ianna, P.
Astronomy Quarterly, 1, 85
1977

Pretty much the same study and results as the previous reference. Additionally, no correlation was found between occupation, medical problems, etc. and angular separation (along the ecliptic) of planet pairs at time of birth.

Dean, G.
Does Astrology Need to be True?  Part 1: A Look at the Real Thing
Skeptical Inquirer, 11, 166
1987

Astrologers prepared horoscopes for subjects correct natal data. Reversed charts were then constructed from the correct charts by retaining the sun sign, but reversing all of the planetary aspects. Half of the subjects were given correct charts, the other half were given reversed charts.  There was no correlation between the perceived accuracy of the charts and whether the subject was given a correct or reversed chart.

Dwyer T.
Unpublished word described in Dean, 1987.

Horoscopes were prepared for correct natal data and for a birth date 5 years and 6 months before the correct date, with the correct sun sign retained.  Thirty subjects were given the correct and incorrect charts.  Half of the subjects picked the correct chart, half chose the incorrect chart.

McGrew, John H., McFall, Richard M.
A Scientific Inquiry Into the Validity of Astrology
Journal of Scientific Exploration, 4, 75-83 1990

Six expert astrologers independently attempted to match 23 astrological birth charts to the corresponding case files of 4 male and 19 female volunteers. Case files contained information on the volunteers' life histories, full-face and profile photographs, and test profiles from the Strong-Campbell Vocational Interest Blank and the Cattell 16-P.F. Personality Inventory. Astrologers did no better than chance or than a nonastrologer control subject at matching the birth charts to the personal data; this result was independent of astrologers' confidence ratings for their predicted matches. Astrologers also failed to agree with one another's predictions.

Geoffrey Dean

Journal of Consciousness Studies, August 2003 (Newspaper  article)

A study of 2,000 people, most born within minutes of one another, tracked over several decades. The study looked at more than 100 different characteristics, ranging from IQ to ability in art and sport, from anxiety levels to sociability and occupation - all of which astrologers claim are influenced by heavenly bodies. He found no evidence of the similarities that astrologers would have predicted.

The experimenters found no evidence of similarities between the "time twins". "The test conditions could hardly have been more conducive to success... but the results are uniformly negative," the research report said.

Dean said that the consistency of the findings weighed heavily against astrology: "It has no acceptable mechanism, its principles are invalid and it has failed hundreds of tests. But no hint of these problems will be found in astrology books which, in effect, are exercises in deception," he said.

Dean is ready for a torrent of criticism: "I'm probably the most hated person in astrology because I'm regarded as a turncoat."  The research undermined the claims of astrologers, who typically work with birth data far less precise than that used in the study.  Dean said: "They sometimes argue that times of birth just a minute apart can make all the difference by altering what they call the 'house cusps'," he said. "But in their work, they are happy to take whatever time they can get from a client."

In 1971 the Survey Research Centre of the University of California Berkeley sampled 1000 adults in the bay area getting information on natal signs and lots of attributes claimed by astrology to correspond. For instance, Leos are supposed to have good leadership qualities. An analysis by Ralph Bastedo found no correlation for leadership, political stand, intelligence, belief in astrology, musical ability, artistic ability, confidence, creativity, occupation, religion, ability to make friends, to organize or to feel deeply. This showed that these tendencies do not differ between signs, so natal signs cannot be used to predict personality traits.

In 1979 Michel Gauquelin put an advertisement in Ici-Paris offering a free horoscope. Recipients were asked to reply saying how accurate they and their friends found the horoscope. Of the first 150 replies, 94% percent said it was accurate as did 90% of their friends and family. Unfortunately, they all got the same horoscope, that of Dr. Petiot, a notorious mass murderer.

In 1985, Harry Edwards checked all the predictions from Old Moore's Almanack for 1984. These were written by a couple of top astrologers. Of the 200 predictions it was possible to check, less than 5% materialized and practically all of those could have been based solely on probability, prior knowledge or astute speculation. Astrologers are no better than pastry cooks, taxi drivers or any of us at predicting.

On June 7, 1989 on American television, James Randi offered $100,000 to any psychic or Astrologer who could prove the truth of their claims. An astrologer who took up the challenge was given the birth information of twelve people and had cast their charts. He interviewed the twelve without knowing who was who and was to identify them by matching them with horoscopes. He got none right. (Note: Randi now offers a $1 Million prize.)

Finally, Michel Gauquelin created a long running affair, starting in the late 60s, called the Mars effect with some research that looked very promising for astrology.

He collected enormous amounts of data from catalogues of famous people ending up with data for thousands of sporting champions, scientists, actors and writers. He found statistically significant correlations between the birth of sports champions and the position of Mars, between actors and Jupiter, between scientists and Saturn and between journalists and Moon. This shouldn't have been too much encouragement because it was only four planets, only four professions, and only for the top couple of percent of those professions. He found no correlations at all for any planet with random samples.

Because of the enormous amount of data it was difficult to reproduce and one attempt at refuting it boiled into a scandal where doubt was cast on the credibility of the refuters. Eventually, a French group, with Gauquelin's cooperation, set up another test of over a thousand sports champions and found a negative result. Gauquelin then argued that some of the champions weren't champions at all and that several other champions had been missed and should have been included. Needless to say these changes produced a positive result but are clearly post-hoc data manipulation introducing bias.

Since none of these tests are positive for a random sample of people it depended exclusively on how you define champion. Gauquelin wasn't cheating or insincere but he was giving himself permission to select his data by focusing on champions and it was just too much work to find nothing.

Edited to add:

Fichten, Catherine S. Sunerton, Betty.

Journal of Psychology. Vol 114(1), 123-134

May 1983

Investigated (1) individual differences associated with horoscope reading habits, (2) the reliability and validity of daily and monthly horoscope forecasts and astrologically based personality descriptions, and (3) the effects of knowing the zodiac sign on the perception of the usefulness of horoscope forecasts and on the accuracy of astrologically based personality descriptions. 366 undergraduates were administered the Eysenck Personality Inventory (EPI) and Rotter's Internal-External Locus of Control Scale. Results indicate that females were more likely to read their horoscopes. Although locus of control was unrelated to horoscope reading habits, Neuroticism on the EPI was closely related. Daily and monthly forecasts were unreliable and invalid. Astrologically based personality descriptions had some reliability, and knowledge of zodiac sign affected ratings of horoscope usefulness and accuracy of astrologically based personality descriptions. (23 ref) ((c) 1997 APA/PsycINFO, all rights reserved)

Peter Hartmanna, Martin Reuterb and Helmuth Nyborga

Individual Differences Research Unit (IDRU), University of Aarhus, Department of Psychology

1 November 2005.

We investigated the relationship between date of birth and individual differences in personality and intelligence in two large samples. The first sample consisted of 4000+ middle-aged male subjects from the Vietnam Experience Study; personality was measured by the MMPI items converted to EPQ (scales) and a large battery of cognitive tests were factored to derive general intelligence, g. The second sample consisted of 11,000+ young adults from the National Longitudinal Study of Youth from 1979. g was extracted from the ten subtests of the Armed Services Vocational Aptitude Battery.

In no cases did date of birth relate to individual differences in personality or general intelligence.

A further goal was to test Eysenck’s notion of possible relationships between date of birth and the popular Sun Signs in astrology. No support could be found for such associations.
We conclude that the present large-scale study provides no evidence for the existence of relevant relationships between date of birth and individual differences in personality and general intelligence.
(Hat tip to Hokum-Balderdash Assay.)

Comments

[Paul]

Hello Richard,
I enjoy your blog, and was especially delighted with this entry, as it speaks directly to an argument I recently had with someone. I have linked to this article from an entry in my blog here: http://journals.aol.ca/plittle/AuroraWalkingVacation/entries/971
I hope that is ok with you.
-Paul

[Skeptico]

Hi Paul:

Glad you liked the article. Please link all you want.

Btw, I liked the article on your site about “spirit photography”. I’ll write a brief post about this later and link to your photos.

[Paul]

Kewl!

[lambic]

Excellent summary of astrology research. In a way it's sad that so much time has been spent investigating nonsense like this, and it's even sadder that astrologers won't listen.

I know of one study that you didn't mention. The abstract is here.

This study was abused by some astrologers who claimed it vindicated astrology when of course it does nothing of the kind. What it showed is that people are willing to incorporate astrological predictions of personality into their concept of self. They also found that daily and monthly forecasts have no validity.

[Skeptico]

Thanks – I’ll add it to my list.

I’ve noticed that people use their supposed astrological “traits” to excuse bad behavior (“I can never make a decision because I’m a Libra” etc), instead of trying to change. Their personality is to an extent formed around what astrology tells them it should be. It’s one of the reasons people believe astrology works.

[buddy]

what an unpopular blog -- I make the 5th person here. WOW

[Skeptico]

Er, check the site meter in the left hand colunm Einstein.

[Chris Mitchell]

It seems that the "Skeptical Inquirer" is as guilty of bad science as the astrologers it seeks to disprove.

It is very true that Michel Gauquelin failed to find any connection between sports champions and Sun, Moon or Ascendant sign. He did, however, and completely against his expectations, discover a positive correlation between sports champions and Mars on an angle (Ascendant, Descendant, MC or IC).

The question is still open as to how significant this correlation is - and there are good arguments on both sides of the debate. However, to conveniently ignore this piece of evidence does no service to science.

The word "sceptic" correctly means someone who is open minded. Sadly, it is often used to describe a mindset that has a particular agenda, and means "anti" (I live in the UK - and the term "Eurosceptic" here doesn't mean someone who is open minded about the European Union, but someone who is strongly against it; the Skeptical Inquirer seem to have the same concept of the word).

[Rockstar]

Chris, I honestly hope you are one of the "fence-sitters" (the people we are trying to help) when it comes to the paranormal, so please excuse my blunt approach.

Correlation proves shit, and does not prove causation. If I found a group of 50 people with severe back pain, and they all had a quarter in their pocket, would that prove anything? That's why real scientists, armed with this and the knowledge of selection bias, shoehorning, and especially confirmation bias pay no attention to this person's "correlations". I'd be willing to wager that if you took every sports champion ever born, no more than chance would be born under a Mars angle. Heck, research it yourself, you could win a million bucks. I read some of the "data" on this case, and would love to show you how easy it is to take a group of people and find a commonality, similar to my brief example. I bet Skeptico already has...

Again, hopefully you haven't made up your mind. Read some more articles on this site and click the links. There are different kinds of skeptics, but readers here are scientific skeptics. We ask for proof, evidence; stuff that is not full of fallacious information. Astrology tried, according to this article, 36 or more times. Let's put it to rest and just say it doesn't work.

[Skeptico]

As you point out, here was considerable doubt about the significance of the correlation, especially since the selection criteria for the “champions” to be included, and their actual correct time of birth, were both in doubt. The result has never been replicated (in fact I believe similar studies in other countries found no correlation), and IIRC the original data have been lost. It’s true that CSICOP made a mess of investigating it, but that doesn’t validate the study, nor does it invalidate the numerous other studies that show astrology doesn’t work.

Don’t you think it’s rather pathetic that after all these years and many studies, the only thing believers in astrology can point to is a dubious, non-replicated study relating to a small aspect (sports stars and the position of mars) of this supposedly all encompassing, detailed and specific fortune telling system?

[Curious Chic]

Astrological Predictions -

Let's see if there is anything to this astrology stuff. Being curious I purchased ($2.00) an astrology program at a garage sale.
The program made two predictions.
1) During the last two weeks of November 2005 the U.S. economy, including the stock market and housing market, is going to experience a MAJOR and lasting and severe (many months) downward spiral.
So, if I believed this I would houseclean my stock portfolio NOW and sell off any stock that I have held for one year or more. I would then repurchase stock after the last two weeks in November 2005 at a great savings.
2) Some of U.S. Vice President Dick Cheney's DARK SECRETS regarding the CIA leak and/or the falsified Weapons of Mass Destruction fabrications will come to light on December 10th through December 16th 2005.

Let's use these two astrological predictions to test this astrological stuff and see if I got my monies worth ($2.00) at the garage sale.

[Cari Lingstrom]

At a young age I began asking people their signs and that of their spouses/friends. There is a definite pattern, as in same elements attracting, and also fire/air and water/earth matches. Also the people I gravitate toward are mainly, though not exclusively, of particular signs. I've come across many people who can't consider astrology in any way as though it is a threat of some kind to even look at it at all. I have not studied astrology beyond what I have described here so have no opinion as to the validity of it's use in predictions, though I do find the idea to be extremely disconcerting at the least. Of course, horoscopes found in the newspaper or written per sign, are absolutely ridiculous and are only for fun.
As for the studies noted above, I find them very interesting though no one will ever convince me there is nothing to astrology. To me it is quite obvious. I see no point to arguing the matter, as it's my experience that there are some who are open to it and others who are just completely blind to it, for whatever reason, choice or personality. I enjoy guessing peoples signs after getting to know them over a period of weeks or sometimes longer. Usually I narrow it down to 2-3 signs, but often I have guessed the first time. I do see that astrology would be very difficult to prove as a science, with all the factors to consider including life experiences. I would never have expected some of the studies described to be able to prove anything. Whether a science or not, it is not a belief, it's real.

[Skeptico]

Carl wrote:

"no one will ever convince me there is nothing to astrology"

There we have demonstrated the mentality of the closed minded believer: nothing could ever convince him astrology is nonsense. Nothing!

Carl, astrology is nonsense you are just completely blind to it, for whatever reason, choice or personality.

[Leon van den Berg]

I blindly believed that medical research was not nonsense and that science proved that low-fat diets prevented heart disease and stroke. On 8 February 2006 The New York Times reports: Low-Fat Diet Does Not Cut Health Risks, Study Finds.

Now this kind of study goes on and on contradicting each other about healthy diets. Shall I conclude that medical research is at the same level as astrology? Seems like it.

[BronzeDog]

I wouldn't trust the New York Times. Read the study itself. The media loves to take any dietary studies and distort them into an excuse for people to keep their unhealthy lifestyles.

Once read somewhere on www.stats.org that one newspaper took a creative interpretation of a study to say that healthy diets don't prevent certain problems. Well, they neglected to mention that the risk did go down, but they took an absolutist view of the word "prevent" to mean zero probability. Since the healthy lifestyle wasn't perfect (just merely better), they decided it was worthless, the exact opposite of the study's conclusions.

[Skeptico]

Leon:

Re: Now this kind of study goes on and on contradicting each other about healthy diets. Shall I conclude that medical research is at the same level as astrology? Seems like it.

No. For one – read the study. It does not say eating a high fat diet is OK.

More importantly, this is just a lame fallacious appeal to “science was wrong before”: just a smoke screen to disguise the fact that there is no evidence astrology works (more correctly – there is virtual proof it does not work). It does not follow that science should not be applied to evaluate claims of astrology, or that nonsense like astrology is likely to be true.

In addition, we know astrology’s provenance: it was just made up. Things “made up” are unlikely to be true. Unlike high fat diets, where we have a tested mechanism that explain why they might be bad.

I am closing this post to comments. In nearly a year no one has come up with any valid criticisms of this post, nor presented any valid evidence that there is anything to astrology, just the same old “be open minded”, or “science doesn’t know everything” lame rationalizations. If anyone thinks they have some astonishing insight into this subject that I have missed they may click the link in the left hand column and email me. If it contains anything of interest I may publish it. Try and make it of a higher standard than Leon's comment though or you may wish you hadn’t bothered.