For a correctly implemented study, what sample size decision should the researchers have made at the interim stage with observed variance estimate being the same as the expected variance estimate? (increase, decrease, or keep the same sample size as the original estimate)