You will read a paper, and the author will find variable X significant at 0.05, and he will go ahead and write his conclusion as if he is 100 percent certain that the effect he found is real.But, as I have just argued, there is a pretty good chance it is not.