Homework Problem 24, Due Wednesday April 28.
-------------------------------------------

Create a 53x4 data-frame exactly like Basstmp from the in-class 
handout on Factors, with Alk as a factor variable, but OMITTING
pH. (The handout can be found under "Factor.txt" under the Handouts
section of the course web-page.) Thus your dataset will have the 
variables 

      logAvM  Alk  Calc agedat   (with Alk a 3-level factor)

(a) Do a linear-model analysis (as in the class-handout) with 
logAvM as response variable, both with "treatment" and "Helmert"
contrasts, and explain the relationships between the two sets of
fitted coefficients.

(b) Do the same linear regression in SAS, using Alk as a categorical
("class") variable, possibly using PROC GLM, and try to recognize 
exactly how the coding of the Alk factor is done by default in SAS
by matching your coefficients with the Splus results.

(c) Using the results of either of your fitted linear models in (a),
give the studentized coefficient that you would use to test the
significance of the coefficient  

    coef[1] - 3*coef[2] + coef[3]

(where coef[1] is the Int coefficient, and coef[2], coef[3] are
respectively the AlkB and AlkC coefficients).