Summary: Collected Notes on Solutions to SPlus
P.M.E.Altham, Statistical Laboratory, University of Cambridge.
December 12, 2006
Not to be quoted without acknowledgement, please.
I originally constructed these draft solutions for our MPhil students in about
1998. I tended to make very limited use of graphics at that time, so now I have
more leisure I will add the odd extra graph here and there, and maybe expand
the explanation in places.
Notes on solution to worksheet 1.
First, Figure 1 shows the plot of y against x. You will see that although y is
certainly an increasing function of x, this dependence is just possibly quadratic
rather than linear. Here is the linear regression (the output is slightly edited to
Call: lm(formula = y ~ x)
Value Std. Error t value Pr(>|t|)
(Intercept) 9.4543 0.6553 14.4283 0.0000
x 1.6579 0.0753 22.0139 0.0000