Íslenskar landbúnaðarrannsóknir - 01.03.1970, Síða 83

COLOUR INHERITANCE IN ICELANDIC SHEEP 81
TABLE 40
Summary af ^2’-5 from Table 39
Matings Singles Twins S + T Heterog.
DF X2 DF X2 DF X2 DF X2
A 12 20.871 12 14.370 12 21.403* 12 13.838
B 12 15.523 12 22.941* 12 24.654* 12 13.810
C 12 22.790* 12 8.229 12 8.588 12 22.431*
B + C-A 12 17.144 12 16.800 12 11.838 12 22.106*
A = all matings
B = matings involving only genotypes AiAi, A1A5 ancl A5A5
C = all matings except those in B
*) P < 0.05
TABLE 41
Comparison of x2’s from Table 39, B
Singles Twins S + T Heterog.
Matings
DF X2 DF X2 DF X2 DF X2
White .... 7 5.546 7 12.998 7 13.366 7 5.178
Nonwhite . 5 9.977 5 9.943 5 11.288* 5 8.632
Sum 12 15.523 12 22.941* 12 24.654* 12 13.810
■) P < 0.05
shown in table 42 as function of the three
parameters u, v, and s. Table 43 shows the
probabilities for white and nonwhite male
and female progeny for the sarne matings,
obtained by summation of the individual
probabilities in table 42.
The observed frequencies in table 39 B
were assumed to be multinomially distri-
buted with probabilities in each class as
shown in table 43. The likelihood was then
of the form
eL oc llpini,
where p^ refers to individual probabilities
in table 43 and n; to the corresponding
frequencies in table 39 B. Tlie log-likeli-
hood function, L oc Snilogp, was used in
the estimation procedure, and the estima-
tion carried out by a general maximum like-
lihood program which finds an iterative
solution to the parameters and their vari-
ance-covariance matrix by numerical differ-