speaker@ttidcb.UUCP (Kenneth Speaker) (01/09/86)
Someone on this net (Craig?) mentioned a controlled experiment where astrologers (astrologists?) attempted to predict personality traits of a set of individuals. The results discredited the predictive abilities of astrology. I would be very interested in reading this paper. Could whoever mentioned this experiment mail or post a reference? Thanks. --Kne
wallace@ucbvax.BERKELEY.EDU (David E. Wallace) (01/16/86)
In article <614@ttidcb.UUCP> speaker@ttidcb.UUCP (Kenneth Speaker) writes: >Someone on this net (Craig?) mentioned a controlled experiment where >astrologers (astrologists?) attempted to predict personality traits of >a set of individuals. The results discredited the predictive abilities >of astrology. I would be very interested in reading this paper. Could >whoever mentioned this experiment mail or post a reference? I'm not the original poster, but I do know the article you're referring to. It's "A double-blind test of astrology," by Shawn Carlson, Nature, Vol. 318, December 5, 1985, pp. 419-425. The importance of this experiment is that it was designed with the assistance of respected members of both the scientific and the astrological communities as a valid, scientific test of astrology as it is practiced today by professional astrologers. It was designed to test what the author calls the "fundamental thesis of natal astrology," that "the positions of the `planets' (all planets, the Sun and Moon, plus other objects defined by astrologers) at the moment of birth can be used to determine the subject's general personality traits and tendencies in temperment and behavior, and to indicate the major issues which the subject is likely to encounter." The experiment described actually consisted of two parts, of which the second part yielded the results Ken mentions. In this part of the experiment, astrologers attempted to match an individual's horoscope with one of three personality profiles from the California Personality Inventory, a widely-respected personality test which was chosen by the advising astrologers as best measuring the types of personality traits discernable with astrology. One of the profiles was of the individual whose horoscope was to be matched; the other two were those of two other individuals, chosen at random. The astrologers predicted that they would be able to identify the correct profile at least half the time; this was the "astrological hypothesis," to be tested against the "scientific hypothesis," which assumed they would only be able to guess the correct profile with a probability of 1/3. The experimenters decided in advance that they would only reject a hypothesis if the results differed by more than 2.5 standard deviations from the prediction under that hypothesis. (The astrologers were asked to rank their choices 1, 2, and 3, and to weight each choice from 0 to 10 based on how good a match they thought the profile was to the given horoscope; this allowed more detailed analysis of the results, described in the article, but the primary test was based on their first choices, as described above.) Out of 116 first place choices, the astrologers chose the correct profile 40 times: 0.256 standard deviations away from the chance prediction of 38.5, and 3.34 standard deviations away from the astrologers' prediction of 58.5. These results were consistent with the scientific hypothesis; inconsistent with the astrological hypothesis. Analysis of the second and third place choices was also consistent with random chance; analysis of the weights showed that there was no significant tendency for the astrologers to be correct more often on the profiles that were rated as strong matches for a given chart (in fact, the chart presented indicates that the astrologers were actually correct less often on those first-place choices they rated as 8-10 than they were on those they rated 4-6). More details on the results and the care taken in the design of the experiment to avoid introducing systematic biases in favor of either hypothesis are presented in the article itself; the above is just my summary of the highlights. I'm cross-posting this to net.origins because I think this article serves as a good example of the willingness of some scientists to investigate rigorously even extraordinary claims, ** provided that such claims can be expressed in the form of a testable hypothesis which yields predictions different from those that would be expected under prevailing scientific theories **. The astrologers were able to bring their claims within the domain of science by casting them in such a form, and were given a reasonable opportunity to demonstrate these claims. They failed the test, but had they succeeded, I believe (and I think most scientists out there would agree) that they would have been taken very seriously indeed by the scientific community. If and when creationists are able to cast their claims in such a form, they too will merit serious scientific scrutiny. Further discussion of this point should go to net.origins only. I don't believe that this article made the pages of Nature and got the attention it has solely because it discredits astrology: most such articles would probably be considered irrelevant by the editors. Rather, I believe that it merits inclusion in those pages because of the the care the experimenters took to ensure that what they were testing was astrologers' concept of astrology, not just scientists' concept of astrology, and the rigor with which they ensured that the results would be scientifically valid, regardless of the outcome. Dave Wallace (...!ucbvax!wallace wallace@ucbkim.berkeley.edu)