Should research success be predicted amongst medical students and junior doctors?