For #7, I don't understand, is the procedure not appropriate because it violated the less than 10% rule? :x
For #8, I reject the null hypothesis because the critical value is less than the alpha level of significance? It's two-tailed, right?
For this one, I would reject the null hypothesis of no difference, right? Because the p-value is low enough, right?
For #14, I would try two types of test (uh, two-sample t-test, and some other one...)
Other than that, what would you say is a better test? IMO, a paired t-test, the difference being L1-L2...
#15, I chose "none" because I believe the problem indicates a correlation. I think.
#16, would it be mean 1 difference ?
For #20, increasing the significance level will increase the power of the test?