
Babies Regression Model
Sorry about the shabby layout of the following question.... im not too good with LaTex Math :).... the data should be in table form, so the first number would be when i=1, second when 1=2 and so on....
The following table shows the results of a study on breast feeding in premature babies. Here the data are observations
on yi (the number breast feeding in group i, i = 1, . . . ,6); on the number not breast feeding in each group
(mi−yi, i = 1, . . . ,6), and also on xi, i = 1, . . . ,6, the gestational age of all the infants in that group (in weeks) at the
time of birth. Note that the xi values here have been used as the basis for forming the six groups.
xi (weeks) 28 29 30 31 32 33
(yi )no. breast feeding 2 2 7 7 16 14
no. not breast feeding 4 3 2 2 4 1
(a) Propose an appropriate logistic regression model with linear predictor b0+b1xi.
(b) Write down the log likelihood in terms of b0 and b1 based on the model in (a)