Methodological investigations of the Hamilton Anxiety Scale.

In an open field study 4184 outpatients were assessed by 508 independent doctors using the Hamilton Anxiety Scale (HAMA). The results of the raw score distributions and item and factor analyses showed that the HAMA is unsuitable for differential diagnosis and, while showing a satisfactory degree of consistency, has relatively poor resolution. Departing from Hamilton's two-factor solution (1959), comparison of various solution models showed that a three-factor solution yields a stable and clinically useful item classification, which is also very similar to Hamilton's three-factor solution (1969).