In: Statistics and Probability
Explain, as if to a friend with no stats experience, why it's dangerous to reach dramatically different conclusions for p-values of 0.04999 and 0.050001 obtained from the same dataset. This is intended to get you thinking about the things that affect p-values, which are some of the same things that affect the width of a confidence interval.