Refer to Case Study 1.6, comparing heart attack rates for men who had taken aspirin or a placebo. Suppose that the observed proportions of .017 and .0094 are actually the correct population proportions who would have heart attacks with a placebo and with aspirin. The following Minitab output shows the power of a one-sided test for two proportions for this situation for three sample sizes. The samples are the number of participants in each group (aspirin and placebo).

Suppose you are the statistician advising a research team about conducting a new study to confirm the results of the old study. The researchers comment that samples of size 500 in each condition should be sufficient, since the effect is obviously so strong, based on the small p-value for the previous study. What would you advise? Explain