Null hypothesis significance testing: a review of an old and continuing controversy