*
* Example 5.2.2 starting page 112
*
open data andy.dat
data(format=free,org=columns) 1 75 sales price ads
print
table
linreg sales
# constant price ads
*
prj(xvector=||1.0,5.5,1.2||)
disp "Predicted Sales at P=5.5, A=1.2" %prjfit
*
linreg(vcv) sales
# constant price ads
*
* Interval estimation in section 5.4
*
@regconfidence(conf=.95)
*
* Test examples from page 121. Note that these tests can simply be read off the
* regression output. The significance levels of the t-statistics in either the
* regression output, or in the output from TEST are all for two-tailed tests. If
* you're interested in a one-tailed test, just divide the significance level by 2.
* So a two-tailed .10 is a one-tailed .05 (assuming the test statistic is in the
* correct tail to reject).
*
test
# 2
# 0.0
test
# 3
# 0.0
*
* Test examples from page 122
*
test
# 2
# 0.0
test
# 3
# 1.0