B is correct. For a one-sided test with alternative hypothesis of the form $H_1: \mu > \mu_0$, the confidence interval can be constructed as $[\mu_0 - (Z_\alpha) * (\sigma/\sqrt{n}), \infty] = [1.2\% - (1.645) * (0.05/\sqrt{36}), \infty] = [-0.17\%, \infty]$

A is incorrect. This is close to the confidence interval associated with the alternative hypothesis $H_1: \mu < \mu_0$, given by $[-\infty, 1.2\% + 1.645 * (0.05/\sqrt{36})]$ where 1.645 is the corresponding critical value for a normal distribution with $\alpha = .05$ on one side.

C is incorrect. The alternative hypothesis calls for a one-sided test; this confidence interval is the result when a two-sided test is conducted.

D is incorrect. The alternative hypothesis calls for a one-sided test. Furthermore, this results from the confidence interval $[\mu, \mu + (Z_\alpha) * (\sigma/\sqrt{n})]$ which is not used in any hypothesis test.

Learning Objective: Understand how a hypothesis test and a confidence interval are related.

Reference: Global Association of Risk Professionals. Quantitative Analysis. New York, NY: Pearson