There is actually a very easy and very accurate way to empirically test one aspect of astrology.
This is the aspect that states that specific astrological signs have a tendency to some specific amount of body mass. Some signs are supposed to be tall, others short. Some skinny, some fat.
This is easily explored empirically using a very large database like the Department of Motor Vehicles driver license database. The most accurate index to use would be simple Body Mass Index (BMI). BMI should cut across the fact that some ethnic groups or races are taller or shorter than others. BMI would not really care about actual height.
What Astrology REALLY is: In essence, Astrology can be defined as the study of the effect of the forces of the universe upon the individual. This is not exactly an absurd notion. The Milky Way comprises millions of stars and planets, and each of them has no more free will than a cog in a wheel. Why then should a human being have so much free will?
The entire planet Earth is a tiny spec of dust within the Milky Way. And each of us humans is even less of a spec of dust on the face of the planet Earth. So why then is it so illogical that humans would be affected by the forces and energies of the universe?
TRUE Astrology may have absolutely nothing to do with Sun Signs and Moon Signs and Rising Signs and Houses, etc. These are just old tools that people had to make due with before the advent of computers.
Computerization has brought about a MASSIVE change in the way that Astrology can be examined. It is now a snap to analyze thousands or even millions of people. For example, a researcher can use drivers license databases that have millions of people in them. Each driver license provides a birthdate, a height and a weight. It also specifies whether eye-glasses are required. This is more than enough information to PROVE astrology categorically, IF there is anything to prove.
Astrology software creates the horoscopes automatically, and data analysis software can accumulate and analyze the data. This analysis can even include multiple regressions.
A FIRST STEP would be see if there are any correlations at ANY point in time according to birthdate. If yes, it must also be REPEATABLE. Indeed, this is as much an INVENTING of Astrology as it is a proof.
Since the proof is being done on unambigous PHYSICAL characteristics, like BMI, height and weight, there is no Barnum effect. If the results are significant, they should be repeatable using a database in another state or another country.
WARNING: You CAN NOT DISPROVE Astrology using this approach. Its failure to detect any statistically significant information could simply be because the researcher is looking in the wrong place. However, you CAN prove it, if there is anything to prove.
I am constantly hearing professors and researchers blathering that Astrology is UNTRUE. How the heck do THEY know? They have NOT approached it empirically. They are just expressing their opinions. The opinions of True Non-believers is just as dumb as the opinions of True Believers.
Draft 1: 3/1/2012
There is no necessity for sun signs and 12 houses, etc. Disproving their validity does NOT by any means disprove that there is such a thing as Astrology.