Hysterics and dysthymics as criterion groups in the measure of introversion-extraversion: a rejoinder to Eysenck's reply.