88 4 
Journal of Agricultural Research 
Vol. XXVII, No, II 
YUEE'S COEFFICIENT OF ASSOCIATION 
The formula for Q is: 
Q AbXaB-AB Xab 
y~Ab XaB+ABXab 
where AB, Ab, aB , and ab represent the numbers of individuals in the 
four zygotic classes of an F 2 population. The probable error of Q is: 
By reference to a table giving values of Q for the various values of p, 
the value of p corresponding to the observed Q is assumed to be the p 
of the observed population. 
A short table of the values of Q for integral gametic ratios where domi¬ 
nants are linked was given by the author (3). It was pointed out by 
Bridges (1) that a different table was needed when dominant and reces¬ 
sive are linked and for accurate determinations it is, of course, necessary 
to resort to interpolations. These difficulties may be obviated by using 
the following formula: 
p 2 
V4-20(0-l)-2 
( 2 ) 
To make this formula applicable when dominants are linked it is neces¬ 
sary to retain the arrangement of the zygotic classes in the following order, 
p o ' -vj * When the linkage is between dominants and Q 
(AbXaB) +(ABXab) ® ** 
is negative, p will be greater than 0.5 and will represent the noncross¬ 
over group. 
In a back-cross the coefficient of association also may be used to 
measure the degree of linkage and the result is not affected by differential 
0-/+Vr=^ 
death rates. In back-crosses p 
2 Q 
EMERSON'S METHOD 
The second method for evaluating p was proposed by Emerson (4). 
Stated in terms of p and 1 — p instead of the r and s of the original 
formula, and taking AB + ab as the crossover classes, the determination 
becomes 
..(3) 
No formula for the probable error of this method has been suggested, 
but the formula for the probable error of a ratio given under the back- 
cross method should provide a fair measure of the reliability of-results. 
This method has the merit of extreme simplicity, but as Emerson pointed 
out it becomes unreliable when there are wide departures from expected 
Mendelian ratios. The method has been extended by Woodworth (7) 
to include multiple-factor characters. In such cases the use of p and 
1 — p for the crossover and noncrossover gametes instead of r and s 
