Several authors have criticized its lack of description of item content and item scores (15). Hamilton himself stated that the value of the scale “depends entirely on the skill of the interviewer in eliciting the necessary information” (1, p56). Considerable interpretation is necessary for rating items; consistency across settings is questionable (25). Other criticisms involve its heterogeneous factor structure and the mixing of frequency and severity of symptoms. Because it covers several facets of depression, patients with low scores on several symptoms can receive the same scores as patients who score highly on a few symptoms; it is not clear whether these are equivalent in terms of severity (17). This raises questions about the internal structure of the HRSD. Having been developed empirically from reports made by patients, the scale is heterogeneous in content and does not meet the criteria for internal construct validity laid down by the Rasch approach to test development. Whether this is a limitation depends on one’s philosophy; the debate between theoreticians and pragmatists is reminiscent of that surrounding the Health Opinion Survey, another scale from the same era that appears to meet its objectives despite far from perfect psychometric properties.