p < .05 (the really important part of this little phrase) indicates that (if the null hypothesis is true) the probability is less than 5% that on any one test of that hypothesis, the sample and the population averages will differ by that much or more.
