On biases in a study of the effects of occupational and individual factors on the risk of neck trouble.