Study finds rejected advances – but not accepted ones – influence our r (cont)