1.1. Have a look at the models (a) and (b) from Example 1.1.

a) As pointed out in Section 3 in Appendix A, one-sided tests are more meaningful than two-sided when testing model parameters. This is one of several good reasons for offering some thoughts on the likely interrelations between the model variables. Do that, and write down your conclusions.

b) Estimate the two models using SPSS. Interpret and compare the results.

You can download the data: bwght.sav

The variables in the data set are:

faminc                 1988 family income, $1000s

cigtax                   cigarette tax in home state, 1988

cigprice                cigarette price in home state, 1988

bwght                  birth weight, ounces

fatheduc               father's years of education

motheduc             mother's years of education

parity                   Kbirth order of child

male                     =1 if male child

white                    =1 if white

cigs                      cigarettes smoked per day while pregnant

bwghtlbs               birth weight, pounds