After a long wait for confirmation (and a crash course in statistics) I am now able to announce the results of veteran astrology researcher Dr Geoffrey Dean’s analysis of my experiments and attempts to prove the theories in Planetary Types: the Science of Celestial Influence.
Over the next few days and weeks, I will, with his permission, post more details on his analysis and general comments about the book, but I wanted to post the ‘short version’ of the results up front. The bottom line is that two of the three tests detailed in the book did not produce results above those expected by chance, but one did.
The whole exercise has been extremely enlightening and sobering. The immediate upshot is that my nagging suspicions of my own statistical naiveté have been confirmed, but the Science of Celestial Influence still holds plenty of promise.
Although there are many novel and intriguing ideas in my book, the prime mover of the book has always been to establish solid experimental results to back up the theories. I initially contacted Rudolph Smit, the researcher behind the awesome http://rudolfhsmit.nl/index.html site, who in turn contacted colleague and collaborator Dr Geoffrey Dean, who agreed to look over my work.
Dr Geoffrey Dean is co-compiler with Arthur Mather of the book “Recent Advances in Natal Astrology: A Critical Review 1900-1976.” He has conducted large-scale studies of astrology, and has authored or co-authored many critical articles and surveys. He has been a full-time astrologer and was the founding president of the Federation of Australian Astrologers in Western Australia. In 1988 he received a Commemorative Bi-Centennial Award from the Astrological Monthly Review (Australia) for his contributions to astrological research. In 2003 he was elected a CSICOP Fellow for significant contributions to science and skepticism.
In examining my results, Geoffrey faced an awkward problem — I had discarded the birth data for my subjects, so the orb used could not be checked. I had used a nominal orb of 8 degrees, but as Geoffrey points out, the exact orb was necessary for calculating the number of aspects expected by chance, which of course were needed for assessing both my experimental group and my control group.
Fortunately I had listed the planets emphasised by aspect in each chart. Because wide orbs increase the number of aspects, it is possible to estimate the orb by working backwards from the number of aspects. To do this, Geoffrey used a computer program to calculate every possible natal chart between the years 1900 and 1980 (i.e. the same year range as my subjects) at half-hour intervals for various orbs. The orb that gave the same proportion of aspected planets as in my experiment was the effective orb that I used.
When applied to my data, the orb thus obtained indicated that my observed number of hits was essentially no different from the number of hits expected by chance, which was 63%. Previously I had estimated the latter using a control group, which gave about 51%. The difference between 63% and 51% is not large, but it was enough to cast doubt on my findings.
The proposed theories in the Science of Celestial Influence (SCI) may have taken some body blows, but they are still standing, owing to the surprisingly high result for my second test of 33 friends and acquaintances, from which I was able to find 31 with planets in the appropriate positions. Geoffrey called this “a good result,” but because I am the only set of eyes that has been privy to the process, it is one that naturally “would not necessarily hold up in court.” Which means, of course, the test needs immediate replication.
So, the upshot is that SCI requires more tests to have its status as a true science categorically confirmed or rejected, and I will be posting the details on this blog in the next day or two, in the interests of attracting research partners to conduct further tests.
Geoffrey also sent me several pages of subjects who had four planets within 1 degree of the ASC to determine their type and then check if the planets matched. He also sent some pages of idealised physical caricatures that had been associated with the planets to see if they matched to my planetary types. After stating my misgivings that photos alone are not always good enough, especially photocopies of black-and-white photos, and that only four planets were used and only conjunctions to the Ascendant were nominated, I made as many confident determinations as I could, but none elicited any results above those expected by chance.
I would like to thank Geoffrey for the time and effort he put into his thorough analysis, and also for his praise of my efforts to put my ideas to the test, rather than simply making unsupportable claims. He said: “…your efforts to put ideas to the test put most astrologers to shame. If your abilities and industry were able to test further astrological claims, you would be adding valuable work to the existing research base.”
I am sure that future tests of will be of great value to all those interested in seeing whether celestial influences have any solid basis in facts and physics.
My thanks go to fellow writer and former fiction crit-partner, Steven Pemberton – http://www.pembers.net – who helped me understand some of the mathematical mechanics of Geoffrey’s analysis. Steven has an A Level in Pure and Applied Mathematics and a Bachelor of Science in Computer Science from the University of York, and I felt he was eminently qualified to comment on Geoffrey’s work.
August 22, 2011 at 8:47 am
I am in agreement with Dean…astrologers read books and take what they read as being the absolute truth. Most astrology books are based on regurgitation of old books…maybe some new theory thrown in that does not hold water. There is a lot to understanding all of astrology but that does not make it so. Nothing personal but most astrologers are great at expost facto analysis but any predictability you can throw out the window. Astrologers always have an excuse or it just does not bother them that they were completely wrong. They take something like the mars on the MC has more soldiers than expected by chance about 3 percent and claim it proves all of astrology even though the effect is small at best. When one starts testing astrological concepts using control groups it all kind of goes to hell in a waste basket. If nothing else it is reliable as economic predictions. As reliable as history lessons. As reliable as fiction …make it up as you go along. But one thing for sure it is not a science nor will it ever be.