Stat 538, 2005, Homework 10, due November 29 in lecture

Multiple linear regression.

This uses the same data set as homework 9.  You should support your answers 
with appropriate output from Minitab.


1) Show that considered by itself, weight of mother at final prenatal visit 
   is not a significant predictor of child's birth weight.

2) Show that if gestation period and sex of the child are also entered into 
   the model, then weight of mother at final prenatal visit is a significant 
   predictor of child's birth weight.

3) Interpret the coefficients fit in part 2).  Think carefully about how to 
   interpret the coefficient for sex of child.

4) Does the model fit in part 2) appear to give good predictions?  Explain.

5) Minitab reports some unusual observations.  Are they having a serious 
   impact on the model you fit in part 2)?  Explain.