The 2008 statistics: what was studied and what came out of it
In 2008 one of the most extensive statistical studies ever undertaken on astrology was published. It deserves a sober account, because it is overstated in both directions.
The design
What was examined was whether any connection shows up between the sun sign and measurable circumstances of life. The data ran to several million records from official sources — choice of occupation, marriage patterns, illness statistics, accident figures, crime statistics.
The approach was therefore as broad as it could be: not an attempt at interpretation on individual cases, but the question whether, across very large numbers, a departure from chance remains.
The results
In some areas statistically striking departures did show up — in the spread of certain occupations across the sun signs, for instance, and in the tendency to marry between particular signs. The effects were small, but with samples of that size they cannot be put down to chance.
In other areas nothing showed up.
The objections
They are to be taken seriously, and anyone who cites the study should know them:
- The seasonal effect. The sun sign is a statement about the month of birth. It is well documented that the month of birth is bound up with a range of traits — through age at school entry, patterns of infection and nutrition during pregnancy. A departure by sun sign may therefore be a departure by season.
- Multiple testing. Anyone who tests a great many traits against twelve signs will find striking values for statistical reasons alone. What matters is whether the tests were laid down in advance.
- One factor only. What was examined was the Sun. Astrological interpretation rests on planets, houses, aspects and dignities — so the study tests something that astrologers themselves hold to be insufficient.
What follows from it
The last point cuts both ways, and that is where the real value of the matter lies.
Astrologers can dismiss a negative result by pointing out that only the Sun was tested. But then they have to say which test they would accept — and that means putting into words a statement that follows from the whole chart, is fixed in advance and can fail.
Until that happens, the matter is undecided, and neither side can call on this study.
See also: the personal chart — what is calculated in it and what is interpreted.