Daniel believes that in a given conficence interval the researchers should change the confidence level from 98% to 99% so that they can get a more narrow margin with which to "pin down" the true population mean. You disagree with Daniel. What is the basis for your argument against Daniel's suggestion?