BEGINNINGS OF CO-OPERATION 55 



It means the same. Students of statistics have found 

 that when P z= 0.05 or less, that is, when there are 

 fewer than five chances in a hundred of such a thing 

 happening as a result of random sampling or 

 "chance," there is likely to be something significant 

 in such results, the more so the smaller the fraction 

 which P is said to equal. 



We make such tests of our experimental results 

 continually, to find how we are getting on, and I 

 shall give probabilities repeatedly. In doing so it 

 must be remembered that these test the data, not 

 the theory— and that the data may vary significantly 

 for unknown reasons, even when we think we are 

 in full control of the situation; and that because 

 there is only one chance in one hundred, or ten 

 thousand, or a million that a thing may happen by 

 "chance" does not mean that it will never happen 

 through what we call an accident; merely that the 

 chances of its happening so, our evidence being what 

 it is, are on the order of one in one hundred, or ten 

 thousand, or a million. 



I will digress even further into the realm of coinci- 

 dence. A Negro friend of mine spent a summer in 

 Europe and while in Paris visited the art galleries 

 of the Louvre. While there he saw a Negro woman 

 busy looking at pictures and on coming closer dis- 

 covered that she was his own aunt. Neither had any 



