Answer
95% confidence interval: $1.03\ltµ_{placebo}-µ_{Naltrexone}\lt1.17$
Since the interval does not contain the zero, we reject the null hypothesis.
But using a drug to avoid gaining about 1 pound does not seem like a good idea. The results have no practical significance.
Work Step by Step
$H_0:~µ_{Naltrexone}=µ_{placebo}$ versus $H_1:~µ_{Naltrexone}\neµ_{placebo}$
In MINITAB, Select Stat -> Basic Statistics -> 2-Sample t
Select "Summarized data"
For Sample 1 (placebo) enter 72 in "Sample size", 1.9 in "Sample mean" and 0.22 in "Standard deviation".
For Sample 2 (Naltrexone) enter 75 in "Sample size", 0.8 in "Sample mean" and 0.21 in "Standard deviation".
Click in Options. Enter 95 in Confidence level, 0 in Hypothesized difference. In "Alternative hypothesis" select "Difference $\ne$ hypothesized difference". Click OK.
Click OK.
