Our statistical standards may be unrealistic for social science, but there was a petition a while ago to lower the cutoff from 0.05 to 0.001. I think that really would help.