The Alephermes Collaborative

Evidence-based Astrology

Research Methods:

Randomized Trials of Astrological Beliefs

I. Astrological Correlates of Significant Earthquakes Worldwide, 1900-2003

II. A Randomized Trial of Astrological Correlates of Significant Earthquakes Worldwide, 1900-2003

Addendum: Tropical Sun Sign Astrology: The Need for (Soul) Age-Adjustment?
I. Astrological Correlates of Significant Earthquakes Worldwide, 1900-2003

Background: Earthquakes are ideal events to study in astrology because their geographical coordinates and time occurrence are accurately recorded by independent observers, as well as their magnitude. Although such events lack the psychological component associated with traditional astrological interpretation, they may provide an opportunity to assess traditional astrological principles, such as those included in the internally consistent system of astrodynes (also known as cosmodynes).

Methods: We obtained published U.S. Geological Survey data for “significant” earthquakes of magnitude 5 and above that occurred from 1900 through 2003 throughout the world (N=137). [1] We used the technique of astrodynes developed by the Church of Light to assess the relative power, harmony and discord of the signs, houses and planets for each event. [2] The calculations were made with software developed by Astrowin.org. [3] The computer output was calibrated against the original astrodyne software program (HoroScop) developed by the Church of Light.

We defined the power index (PI) as the cosmodyne power of each sign, house (Placidus) and planet (Sun, Moon, Mercury, Venus, Mars, Jupiter, Saturn, Uranus, Neptune and Pluto) relative to the mean cosmodyne power for all signs, houses and planets, respectively. That is, the PI of the Sun for each earthquake was defined as its cosmodyne power divided by the average for all planets, and so on for each planet. We defined the PI likewise for each sign and house. We considered significant a PI > 1.5 or PI <0.7 . No formal statistical tests were conducted in this exploratory study.

Results: More than 95% of all PIs were symmetrically distributed between 0.6 and 1.4. No planetary or zodiacal PI was found significant. The 1^st and the 10^th houses exhibited PIs of 1.9 and 1.7, respectively. The discord indices (DI) were similarly increased for these houses (DIs of 1.7 each). In addition, the harmony index (HI) for Jupiter was significantly increased (HI=1.8), as well as the discord index (DI) for Saturn (DI=1.6). We found no “dose-response effect” (i.e., higher index for higher magnitude) between the earthquake magnitude and any of the astrological indices under consideration in this study.

Conclusions: Earthquake activity seems to be correlated with significantly increased PIs for the 1^st and 10^th houses, as well as with astrological influences that facilitate the expression of Jupiter and afflict the expression of Saturn. These findings are in agreement with traditional astrological principles that consider the cardinal houses as most influential in a horoscope --particularly the 1^st and the 10^th cusps—and the powerful effects of Jupiter and Saturn. Further research is needed to ascertain the interaction of these astrological factors in relation to earthquakes.

*
* * *

II. A Randomized Trial of Astrological Correlates of Significant Earthquakes Worldwide, 1900-2003

Background: We previously reported in the first part of this study that cosmodyne indices may be correlated with earthquake activity. In the second part of this investigation we designed a controlled study to assess the reliability of such results.

Methods: We used the tool provided by Randomizer.org to randomly select months, days, hours and minutes, and substituted these for their corresponding values in the earthquake database previously developed. The year, longitude and latitude remained unchanged.

As previously reported in the first part of this study, we used the technique of cosmodynes developed by the Church of Light to assess the relative power, harmony and discord of the signs, houses and planets for each event. Calculations of cosmodynes were made using software developed by Astrowin.org.

Results: More than 95% of all PIs were symmetrically distributed between 0.6 and 1.4. No planetary or zodiacal PI was found significant. The 1^st and the 10^th houses exhibited PIs of 1.7 and 1.8, respectively. The discord indices (DI) were similarly increased for these houses (DIs of 1.6 and 1.7, respectively). In addition, the harmony index (HI) for Jupiter was significantly increased (HI=1.8), as well as the discord index (DI) for Saturn (DI=1.7). [Note: To further assess the contribution of the variable year in the planetary HI and DI, we subsequently randomized for the year too, and obtained similar results].

Conclusions: Randomization is the preferred method to control bias in the scientific research of events such as earthquakes. Of course, astrological influences cannot be randomized, but a randomized control group, as similar as possible to the study group except for the exposure of interest, is feasible in astrological research. Replication methods to control for demographic factors in astrological research in discussed elsewhere by others.[4]

We found virtually identical results using either earthquake or comparable random data. Our findings suggest that the correlation found in the first part of our study may be due to the astrodyne algorithm itself favoring traditional postulates of house and planetary influences, such as Jupiter given more power and Saturn more discord. However, other traditional postulates of astrology did not appear to influence our findings related to the occurrence of earthquakes.

Regardless of the source of bias, we recommend the use of randomized control groups in conducting astrological research. Also, a holistic approach in testing astrological hypotheses may be needed to replace the current reductionistic assessment of separate components as potential astrological correlates of events. A heliocentric approach may also be needed to study nonpsychological planetary events such as earthquakes.

*
* * *

Tropical Sun Sign Astrology:
The Need for (Soul) Age-Adjustment?

Background: A review of the facts (see Dean-Mather's Invitation on Astrological Research) shows that 1) half the population of Western countries reads sun sign columns at least sometimes, 5% take them seriously and 1% read them often and take them very seriously (Skeptical Inquirer, 1992); 2) well designed observational studies with adequate statistical power have failed to support the tropical sun sign hypothesis that postulates a correlation between personality character traits and natal Sun signs (Recent Advances in Natal Astrology: A Critical Review 1900-1976, 1977); and 3) experimental studies using human subjects are impossible to conduct to test this hypothesis. From this evidence we conclude that: 1) there is a very real and deeply ingrained human need of archetypal proportions to belief in metaphysical causation; and 2) "the sun-sign idea is simply not valid" (The Validity of Astrological Theory, 1984), at least as tested so far.

Alternative paradigm: A few centuries ago the radio wave hypothesis was not falsifiable. Today it is a valid fact. The difference: development of adequate instruments of observation. The intrinsic limitation of the scientific method is that there is never a FINAL and absolute validity determination on any hypothesis because its validity may change with new evidence.

It is possible that the analogy holds true for the tropical Sun sign hypothesis. We may lack the adequate instruments of observation to assess the age of the incarnating human soul (see Alan Leo's Esoteric Astrology). Of course, the theory of reincarnation is itself an unproven hypothesis. However, if the principle of reincarnation operates it would introduce a serious bias not previously considered in Sun sign research. In the same way that international mortality rates comparisons need to be "age-adjusted" for us to draw valid inferences about the health and disease status among countries, tropical Sun sign research design may need to include a soul age-adjustment factor. Younger souls may handle Sun sign energies differently than more mature souls, thus introducing a significant source of bias. On the average no differences would be observed if the analysis is not stratified by this potential confounding factor.

Research proposal: At present we do not know of any reliable and objective way to assess the age of the human soul. Therefore our hypothesis remains unfalsifiable until adequate instruments of observation are developed. However, social class, as used in epidemiological studies in Great Britain, may be considered as an imperfect proxy for the age of the human soul. In general, and with increasingly more pervasive exceptions as we all progress towards more egalitarian societies world-wide, more mature souls may be expected to be born, raised or to live among higher social classes and the reverse is expected to occur with younger ones. Epidemiological research in Great Britain has extensively documented a strong correlation between social class and disease status.

Our proposal consists of conducting an observational study (case-control design) using computerized hospital discharge data of a representative sample of hospitalizations in Great Britain for a given period (the earlier period the better as the correlation of social class with the age of the human soul would be expected to be stronger in the past than in the present or future). Such computerized hospital discharge records should provide, at least, access to the following variables: 1) month/day/year of birth (and of hospitalization, if possible); race/ethnicity/country of origin; gender, occupation/social class proxy, and the International Classification of Disease (ICD) codes for the primary and secondary reasons for hospitalization. It is crucial to identify if several hospitalizations belong to the same individual during the study period (individuals versus events) in the data set. Well defined medical and surgical diagnoses (using ICD) such as diabetes, myocardial infarct, appendectomy,etc. will be used.

Null Hypothesis: Natal tropical Sun signs (the "exposures") are randomly distributed among individuals requiring hospitalization in Great Britain. Therefore, compared with individuals not having Disease Y, the relative risk of Disease Y (as approximated by the exposure odds ratio in a case control study design where the rare disease assumption holds) is 1.0 for individuals born in sun sign X.

Brief Study Protocol: The analyses will stratify by social class and gender, and will be restricted to adults 30 years of age or older. The sun sign distributions will then be further adjusted for demographic (birth cohort) and astronomical (orbit) corrections. Sample size will be limited by availability of computerized hospital discharge records. Assuming specificity of effect, that is, that only one sun sign will be associated with an increased risk of a given disease or surgical condition, and having approximately 5 controls (not having Disease Y but hospitalized for other diseases) per case, a rough sample size of 600 cases and 3,000 controls BY GENDER AND SOCIAL CLASS STRATUM would be needed to find an increased risk of 50% (odds ratio=1.5) with an 80% study power at a significance level (type 1 error) of less than 0.05. The selection bias inherent in using hospitalized controls (see: classical British case-control studies on lung cancer and smoking by Dunn) would be addressed in a further refined study protocol.
Stratified by gender and social class,  |  Disease "Y"  |  Other     |
and restricted by age                   |               |  Diseases  |
----------------------------------------|---------------|------------|
Sun Sign "X"                            |   12%         |    8%      |
All other sun signs                     |   88%         |   92%      |
---------------------------------------------------------------------
                                           100%            100%
Conclusion: We find it both unethical and invalid to continue using Sun sign astrology for commercial purposes. We agree that "testable claims are often made, so their truth or falsity is an empirical matter to be established by research, not proclaimed by fiat or simply ignored." (Astrological Journal, 1996). However, we disagree with the proposition that we may be ever able "to devise a test that will show whether those ideas are right or wrong." It is well known that the result of any statistical test is not a personal conviction but a probabilistic statement of certainty. It is not within the scope of science to "prove" ideas right or wrong. It is within the scope of science to challenge the testable claims originating from those ideas, even more forcefully if such claims are used to misinform the gullible for commercial purposes.

We believe in the psychological causation of disease and consider astrological energies to be part of the individual's psychological makeup. Rigorous epidemiological research in medical astrology may present us the opportunity to examine more objectively defined effects of astrological energies that so far have eluded serious investigators in this field. However, we are concerned about the unethical use of such knowledge, if found valid, by health insurance companies and others. The mapping of the human genome is already presenting such ethical dilemma. Something similar is to be expected if the key to medical astrology (psychogenetics) is discovered.

References:
The Validity of Astrological Theory, PhD Thesis, London University, March 1984, 246.
Skeptical Inquirer 1992, 16, 344-347.
Astrological Journal May/June 1996 pp 143-155.

THOTh, last revised 03NOV96

*
* * *

| Home |

2010-10-23