Informed versus blind: The reliability of cross-sectional ratings of psychopathology