Rarely has anyone actually defended null hypothesis significance testing