If the .05 significance level seems somewhat arbitrary, that’s because it is. There is no single standardized statistical threshold for rejecting a null hypothesis. Both .01 and .1 are also reasonably common thresholds for doing the kind of analysis described above.
