Suppose that Elsa and Frank determine confidence intervals based on the same sample proportion and sample size. Elsa uses a larger confidence level than Frank. How will midpoint and width of confidence intervals compare