As a followup to my question about the birth month of boxers, I am posing the fundamental question along with my hypothesis (testing if there is any truth to a conclusion an astrologer might make): That good professional boxers would tend to be Taurus, Capricorn or Virgo (the 'earth signs', people of stamina, strength, earthiness, tenacity, power and control). And that we should see fewer 'ethereal' signs in pro boxing (that is statistically significantly fewer Pisces, Aquarius and Libra—these tend to be more peaceful, docile, 'high minded', fairer, less brutal, less earthy).
I am specifically using boxers and not, for example, NFL players because many people might play football for different reasons and have different skills (speed is obvious example, or ability to throw a ball far, or ability to read a defense aka intelligence). But a professional boxer (particularly middle weights and above) must have strength, endurance, and power (lighter weights might get by on speed alone). Even very fast heavyweights like Ali are quite powerful and possess enormous determination.
In any event, any statistical difference between population birth date/sign averages and those of boxers would be interesting even if causality could not or would not be established. I might then look at other areas—for example, birth dates of serial killers, birth dates of song writers. Astrologers might say that 'Capricorns make good accountants' but how does one get the birth dates (reliably) for accountants? The birth dates for pro boxers and for serial killers and other targeted populations can be found.
My results are these (as mentioned in my previous question about birth months of boxers—I avoided mentioning astrology at that time because it makes some people moan and/or think you are crazy. I'm just curious and investigating).
n = 67. 27 fit into earth signs (my predicted Capricorn, Taurus or Virgo). Is that statistically significant? Is 27/67 statistically different from the predicted 25% of 67 (16.75)? (There are 4 sign categories so I expect 25% to be in any particular sign category).
I find also that 10/67 are the 'anti-boxers' that I predicted—that is, only 10 of 67 are Pisces, Aquarius or Libra. Again, is this statistically significant compared to the expected 16.75 (1/4 of 67).
Based upon the previous answer to my question, it appears the chi-square test needs to be done, or does my posting here alter the best way to approach this?
I also want to update my list, getting closer to 100. I also may want to refine and separate out flyweights and anything below middleweight—however I think this will reduce the sample size down too far and perhaps it is making a bad assumption about the need for strength, power and stamina in lower weight class fighters.
Does chi-square remain the best way to test this data? Or is there a way to test it using a 'category' approach (12 categories, 1 per sign, or perhaps 4 categories—the first category is the earth sign, the other category is the Pisces/Libra/Aquarius category of 25% of all births)?