A random sample of 800 men and women between the ages of 18 and 64 living in Camberwell, south-east London, were interviewed using a short form of the Present State Examination (PSE). The interviewers were from a professional Agency and had received a shortened version of the usual PSE training course. A further interview was sought one month later with all subjects who were above the threshold on the Index of Definition, and a sample of those below the threshold. At this interview the full PSE was administered by members of the MRC Social Psychiatry Unit. Forty-nine audiotapes of Agency interviews were available for the members of the MRC team to rate.
This paper reports the reliability between interviewers. Techniques of comparison were chosen which enabled interviewer reliability to be assessed, without making the assumption that individual subjects would display the same symptom levels over a period of one month.
There is strong evidence that at least 2 of the 8 Agency interviewers had thresholds that were lower than those of the MRC team for rating many of the common minor symptoms such as irritability, also the key symptom of depressed mood. At least 4 of the Agency interviewers, however, were rating in a similar manner to the MRC team. There was no marked tendency for thresholds to change over the 6-month period of interviewing.