When data is paired, what would be the consequence of conducting two-sample/independent-sample tests and forming confidence intervals assuming the two samples are independent? (choose all answers that apply)

a. We lose power to see differences between the samples.

b. We overestimate the true standard error of the difference in the means.

c. Nothing. It would be fine to use the independent sample procedures. We just have another option as well.

d. We underestimate the true standard error of the difference in the means.

e. We would gain power to see differences between the samples.