Section 9.1 (p.524)![]()
1. 

2. From Exercise 1,
. Rewrite
as follows:


![]()
Therefore, 
4. According Equation (6) or Exercise 2, Q = 8.6. Under the
null hypothesis H0 that the 2 blood types are independent, Q has a c2
distribution with 3 df. Therefore we reject H0 at a0
= 0.05, but not at a0
= 0.025.
5. Similar to Exercise 4, we have Q = 0.91 which has a c2
distribution with 1 df under H0. We reject H0 at a0
= 0.4, but not at a0
= 0.3.
8. Follow the procedure discussed in the book.
(1) The actual number of
unknown parameters when H0 is true is R+C+T-3
(2) Q has the form
where
is the MLE.
(3) When H0 is
true and n ® ¥,
Q converges to a c2
distribution with RCT-R-C-T+2 df