68 



TECHNICAL BULLETIN 7 



JL Correlation Between Times Broody and Annual Rate for Broodi/ Birds 

 Alone — Pullet Year. 



A pure race in so far as the broody trait is concerned is to l)e found in tlie 

 birds actually going broody during tiieir first laying year. The total number 

 of birds in this class for the seven years is 1122. By tabulating the annual 

 rate of eacli individual against her number of broody periods a measure of 

 the degree of correlation between degree of broodiness and annual rate is 

 obtained. Constants calculated on this group follow: — 



Number of i)irds 



Mean times liroody 



Times broody standard deviation 



Mean annual rate 



Annual rate standard deviation 



Coefficient of correlation 



Regression broodiness on rate 



Resrression rate on broodiness 



112-2 

 2.89 

 ±1.91 

 .54.93 

 ±9.24. 

 —.3232 

 —.0669 

 —1.5610 



.0180 



The mean annual rate for the broody birds is slightly lower than was found 

 for the total population in section 11 (56.48). No significant change is observ- 

 able in standard deviation. 



Tlie coefficient of correlation is slightly larger than that obtained for tlie 

 total number of birds, and represents a rather intimate negative correlation 

 between times broody and annual rate. Degree of broodiness as measured 

 by number of periods is therefore very inimical to high annual rate. 



13. Correlation Between the Presence of Broodiness and Annual Rate above 

 the Mean of Broodies and Non-broodies Combined — Pullet Year (Flocks 1910- 

 1923). 



The true relation or correlation between the presence of broodiness and 

 annual rate above the mean is of interest and value to poultrymen. Such a 

 determination has been made for the 2245 birds being studied, by Yule's 

 metliod. 



Coefficient of correlation 



.2828 



.0131 



Tlie above constant does not differ significantly from tJiat representing the 

 whole population. In this particular case the mingling of a broody and a 

 non-broody race in the same correlation table did not result in skew correla- 

 tion. The constant — .2828 ± .0131 is known to represent a true value for 

 tlie flock in question, and emphasizes the importance of breeding for non- 

 broodiness to secure maximum annual records. 



