Questions to ponder:

You are an ecologist investigating whether there is a correlation between anchovy density estimates obtained from otter trawls and western gull abundances.  You run a regression and get a p value returned of 0.867.

Can you reject the null hypothesis?

What would be the probability of committing a Type I error?

If there really is a direct relationship between anchovy density and western gull abundances what type of error have you likely committed?

What can you do to reduce the probability of this kind of error?