Accuracy of behavioural assessment by computer.

An automated assessment interview was given by a microcomputer to 26 randomly selected patients, referred for treatment of phobias. The results were compared with those of conventional clinical assessment by experienced behaviour therapists. Ratings of overall severity and intensity of specific types of agoraphobia and social phobia were derived from the computer interview, and correlated very closely with global ratings by the clinician and also with an independent structured clinical assessment. The computer also elicited behavioural targets to serve as a basis for exposure treatment. Blind assessors rated these targets as highly as those arrived at by clinicians in respect of their practicability, precision and appropriateness for treatment. The automated technique is inexpensive, saves clinicians' time and can be made widely available for screening, assessment and progress monitoring. It may also provide a basis for automated exposure treatment.