If the sample were biased in some way, how would that affect the interpretation of the confidence interval?

a) It would cause the confidence interval to be wider than it should be.
b) It would cause the confidence interval to be narrower than it should be.
c) It would make the confidence interval more accurate.
d) It would not affect the confidence interval.