An astrological hypothesis is the claim that a correspondence between a group of persons (or events) with a certain property, and a certain constellation of their horoscopes, is "not by chance". A very dumb example: Artists frequently have Sun in Libra - which amounts to the (false) claim that there are more artists with Sun in Libra than could be expected by chance.
astrotest assists you in testing astrological hypotheses. For performing such a test, three things are necessary:
- 1. Objectify your hypothesis.
-
For the work with
astrotest, this means: Define a score function which expresses a function built of elementary constellations of a horoscope, like zodiacal or mundane aspects, sign positions, or house positions. Inastrotest, each term consists of a weight factor (which can be negative, converting the term into a penalty term), applied to an elementary constellation like "Sun in Leo" or "Angle Mercury / Venus between 40 and 50 degrees". The full score function is a collection of those score function terms. Define a score function such that your hypothesis can be expressed as "The horoscopes from this group have a score higher than could be expected by chance". - 2. Collect birth data...
-
...for your group with the given trait. Do this diligently, following a strategy. If you don't want to fool yourself, don't look at the horoscopes themselves at this stage! Don't base the decision to add or to ignore a certain horoscope on any astrological criterion! You may use given horoscope collections like Taeger's Internationales Horoskope Lexikon, encyclopedia,... For
astrotest, the horoscope data have to be available in the form of an AAF data collection, or as a CSV file with julian dates, geographical longitude and latitude. - 3. Perform the test run...
-
to check your hypothesis. A p value lower than 5% makes your finding interesting - worth to perform an independent re-check with different data. Such a replication studie with a completely different data set, and still with a good p value, can be regarded as a good validation for your hypothesis.