of essential facts, lies in her ability to learn quickly the requirements of the job. Thus the intelligent interviewer will shine during the training period and bring her interviews quickly to an acceptable standard. But after a while (and interviewing went on over several months) the less intelligent will catch up, and beyond a certain level of proficiency the intelligent interviewer actually has no special advantage.38 There are two additional conditions which diminish the advantage of intellectual ability. First, the original selection procedure narrowed the range of this ability, practically assuring that all interviewers had the minimum required to attain satisfactory standard. Second, the prolonged training period equalized somewhat the different original abilities. It is likely that differences in intelligence would be more important where only little training can be given. Conversely, differences in degree of training may compensate for differences in intellectual ability of interviewers.
The supervisors' ratings encompassed a longer time period than did the editing errors in the first few interviews, but they are not related to the over-all performance as measured by all the questionnaires. As we have seen previously they are related to those interviewer traits which we have called motivation; and the basis of the evaluation may not have been those factors which were used in our performance measures. In fact, these evaluations may supplement such measures. Apart from the obvious consideration that getting along with a supervisor is important in completing a job assignment, it appears from discussion with the supervisors that they took into account a number of criteria not visible in the completed interviews; for example, the number of refusals or interrupted interviews which somebody else had to complete, general perseverance (because of which the interviewer was frequently assigned very difficult cases), effort, contribution to morale of the team, etc. In the long run these accomplishments turn out to be a function of motivation, once initial differences in ability become negligible. The only factor which the supervisor could not take into account was what went on generally in the interviews in terms of communication (although they did sit in on a few interviews). The relative importance of the personality factors and the factors the supervisors did consider cannot be assessed in these studies where the supervisor's duty was to produce the highest quality work and where she could help some interviewers and adjust their assignments. In a purely methodological study with random assignments supervisors could be restricted in their performance and prevented from compensating for specific weaknesses. In this case we would expect more definite relationships between interviewers' traits and performance, but the quality of the whole survey would suffer. Here we can say that motivation and the personality traits which we have defined are both important.
The picture which emerges from the correlational analysis can be illustrated by our experience with individual interviewers. Revisions of early judgments about particular interviewers are especially instructive for this purpose.
Interviewer A had had some experience in this kind of work, enabling her in the screening interview to understand quickly the purpose of the study and to achieve high ratings on intelligence and poise. From the beginning, however, she did not make use of her potentialities, giving the impression that the importance of her background had been overrated. Eventually, rated very low by her supervisors, she also rated low on performance criteria.
In contrast with her was interviewer B, who had one of the lowest educational and experiential backgrounds. She was unimpressive during training and was even considered to be doubtful during the beginning of interviewing. At each periodic review of the interviewers she was rated higher, until she was considered to be one of the best interviewers. Her efforts and perseverance made her valuable for the study during difficult times during the survey. However, her rank on the performance criteria was only average.
These two examples show the importance of motivation, especially in extreme cases. Interviewers who scored high on the performance criteria were more inconspicuous, and their changes in performance were less dramatic. It may be that the personality traits which we have identified characterize interviewers who are neither problems nor outstanding at any stage of the study.
The dilemma of the practical use of the interviewer evaluation becomes apparent from these descriptions. The traits which are most easily identified--those belonging to the pattern of intellectual ability--turn out in the long run to be least important in an extended survey. The variables most difficult to identify and to measure, the personality traits, will affect most of the data obtained in the interview. More definite pinpointing of these factors must await further development of performance criteria of interviewing and studies in the assessment of interviewers. It must be kept in mind, however, that even a partially effective selection procedure, together with extended training and efficient supervision could reduce the variability between interviewers. At this point the variability is still noticeable but sufficiently small so that it does not critically affect the substantive results of the study.
Interviewing is a process of transmitting information, but it also includes personal relations between two people. A well-trained interviewer will establish a personal relationship which is most conducive to the conduct of the interview itself which will be in general quite emo-____________________