The idea here is to have a little bit of scientific fun with astrology, the history of which we will discuss in class on Wednesday. For three days, record in your journal astrological forecasts from two newspapers or other sources of your choice (be sure to state what they are). Also record what happens in your life for those three days. Is there any correlation? Formulate and discuss a hypothesis to explain your results. What tests might you perform to confirm or falsify your hypothesis? Also record your birthday, and record the number of the forecast below which best matches the sort of week (1/27-2/2) you have had (without knowing which forecast is for which sign, of course!). Finally, how many of your colleagues do you expect will pick the ``right'' forecast, and why?
Many people pointed out that most of the forecasts are generally vague enough to apply to almost anyone. One pointed out that you can read through lots of wrong or inapplicable forecasts, but it's the one which happens to be correct that sticks in your mind, making astrology seem valid. A few people also noticed the fact that forecasts from two different sources often seem totally at odds with each other!
A scientific hypothesis is an idea, inspired perhaps by some observed phenomenon, which can (at least in principle) be falsified by some experiment or observation. In the case of astrology, the null hypothesis, supported by all the known evidence, is that astrology doesn't work: natal charts and Sun-sign forecasts have no correlation with human personalities and actual future events.
This assignment is an example of a scientific test of this null
hypothesis: give people a list of forecasts with the signs removed and
see if there is any correlation between their actual sign and the
sign of the forecast which they pick as most applicable to their
lives. If we assume the null hypothesis, people will pick forecasts
entirely at random, and the probability of picking the correct one
will be 1/12. For a class of 25 people, we would expect about
people to pick the right forecast.
Of course, sometimes just by chance in a given sample of students,
fewer or more will pick the right one, but if you try the experiment
lots of times, you expect the average to come out close to 1/12. And
the more times you try it (or the bigger the class size), the closer
you expect to get (the ``law of large numbers'').
Other scientific tests of astrology have been performed, most notably the work of Shawn Carlson (see the article in the reader, p. 58), who carried out a careful double-blind test, with techniques approved by the ``professional'' astrological establishment. No one has found conclusive scientific evidence to support the astrological hypothesis that the heavenly bodies directly influence our lives. Given the highly bizarre and unnatural new ``force'' which would be required if the astrological hypothesis were correct (see Fraknoi's article in the reader, p. 53), it is then natural to discard astrology as a science (Ockham's razor: ``hypotheses should not be multiplied unnecessarily'').
Of course, religious or mystical beliefs are generally outside the realm of scientific inquiry and therefore cannot be criticized scientifically (except when the beliefs are interpreted literally or scientifically). The problem with astrology is that it usually opens itself up to scientific falsification by making predictions about future events or human personalities. It is a natural product of a world view in which humanity and earth occupied a central location in the cosmos, and we now know better!
This document was generated using the LaTeX2HTML translator Version 98.1p1 release (March 2nd, 1998)
Copyright © 1993, 1994, 1995, 1996, 1997, Nikos Drakos, Computer Based Learning Unit, University of Leeds.
The command line arguments were:
latex2html -up_url ../../assign.html -up_title Astronomy 9 Assignments -split 0 -t Astronomy 9 (Spring 2000): Assigment 4 Solution Notes -dir /coma8/jbaker/public_html/courses/ay9/week2/hw4sol hw4sol.tex.
The translation was initiated by jonathan baker on 2000-02-09