Reliability of clinical assessment of blunted affect

The affect of 20 chronic schizophrenic in-patients was assessed by two psychiatrists using two rating scales. The interview recordings were reviewed three times--firstly with the auditory information only (the "audio review"), secondly with the visual information only (the "visual review") and thirdly with both (the "audio-visual review"). The "blunted affect score" correlated strongly with the total score of the verbal items of scales, but did not when a single mode of information was given. Since the absolute values of the "blunted affect score" were lower in the "visual review" than in the live interview and in the "audio-visual review", it was speculated that speech tended to bias the assessment of affect.