Interpretation and Hypothesis Testing in Bivariate Regression

a) SAS/Stata Tasks.

i) Regress bmiR on age .

ii) In SAS, use the /clb option to request the 95% confi dence intervals (recall that Stata calculates the confi dence intervals by default).

b) Write-Up Tasks.

i) Write the null and alternative hypotheses about the slope that are being tested by the t-value and p-value listed in the output. Why is this null hypothesis interesting? Write the steps for making a decision about these hypotheses (use alpha=.05). Interpret the point estimate of the slope.

ii) Calculate by hand, and interpret, a 95% confi dence interval for the slope.

Note whether and how the confi dence interval results relate to the t-test examined in Question 8.1bi.

iii) Repeat the hypothesis test from Question 8.1bi, but use a one-sided alternative, based on the hypothesis that older adults have a higher body mass index. Be sure to write the null and alternative hypotheses and indicate the t-value and p-value for this test.

iv) Write the null and alternative hypotheses about the intercept that are being tested by the t-value and p-value listed in the SAS and Stata output.

You do not need to go through the steps of making a decision about this null hypothesis, but you should comment on whether this test is meaningful in this situation.