Misuse of the Symptom Checklist 90

To the Editor.—  My career-long interest in the complex issues of physicianpatient agreement was one of the instrumental factors in my development of the Psychopathology Rating Scale Series,¹ which includes the Symptom Checklist 90—Revised (SCL-90-R) and SCL-90 Analogue. For this reason, I approached the recent article "Discordance Between the SCL-90 and Therapists' Psychopathology Ratings" (Archives 1983;40:389-393) with enthusiasm. On reading the article, however, it became apparent that the study suffered from serious methodologic flaws, a striking misappreciation of basic psychometrics, and a surprising amount of plain misinformation. As a result of these problems, the more glaring of which are listed below, I have come to the judgment that few (if any) of the study's conclusions are valid.1. The investigators used an early prototype² of the SCL-90-R,¹ not the final published version of the scale. The SCL-90 Analogue and its definitions were developed to provide consonant measurement