We tend to overstate our negative feelings and symptoms in surveys, shows a new study. This bias wears off over time, but the results point to the possibility that measurements of health and well-being, which are vital in making medical assessments and in guiding health-related research, may be misinterpreted.