Whereas the use of statistical procedures is well controlled by generally accepted rules, the use of elaborate personal interviews is less well regulated because they have not always been recognized as valid tools of social research. The few schools which have made use of them (for instance, Chicago) have done so more on a hit-or-miss basis. The Princeton Radio Research Project, directed by psychologists, has a twofold task: to bring the use of this kind of material into the foreground and to deal with it as systematically as possible. Although it is premature to attempt any systematic analysis now, the matter should be discussed among us so that exemplifying material can be collected and the general principles slowly developed. The following remarks tend only to serve as a first step in this direction; they are only loosely interconnected.
1.) Detailed interviews are necessary for the description of psychological processes. If, for instance, we want to know whether people who listen to a radio drama have visual imagery, or if we want to distinguish the emotions the people experience when listening to music, such descriptive interviews are necessary. Their use in this connection goes back to the Würzburg School where they have been used to describe the processes of thought, of solving tasks, etc.
Behind this approach is the assumption that some of those processes are of a very general nature and hardly different for different individuals. Whenever such use is made of them it will have to be pointed out why the student assumes that individual differences can be neglected.
2.) This does not mean, however, that such psychological descriptions cannot lead to generalizations; it only means that we are not interested in the frequency distribution of those individual experiences. In a Viennese study, for instance, it was shown that when people try to match pictures with voices they use one of two procedures: either they make their judgment intuitively or they try to arrive at a judgment by rational deductions from clues; it was shown statistically that the first procedure is much more successful. Similarly, Dr. Wiesengrund thinks that the less emotional a person’s attitude toward music is, the more he knows about it. It is evident that even a few cases should prove or disprove such assumptions. It would be worthwhile to conduct a systematic investigation as to where this kind of statement could be found in the different sections of the project.
3.) It will be useful to distinguish the analysis of effects from such a description of general experiences. The »Ohio School on the Air«, for instance, has sent an elaborate questionnaire to people who had listened to one of their art courses.1 In about fifty questions, they tried to circumscribe the effect of the course on the listener. A more modest effort of this kind has been made in our WIXAL2 Study. There seems to be a growing tendency among our staff members to use this kind of approach: How has listening to the radio influenced your newspaper reading? Do farmers pay fewer visits to their neighbors since they have a radio? Etc.
It is probable that results can be gained this way especially if the questions are very specific. (This type of interview is certainly easier in regard to one special broadcast than in regard to a whole program series, let alone radio listening at large.) However, it is not certain whether the results will be valid without statistical corroboration; there seems to be no justification to assume effects which are alike for all individuals. The best use to be made of this kind of interview is probably to gain leads for further statistical studies on special points.
4.) The most tempting use of the detailed interviews is connected with the question of why people listen to a certain program in the sense of what it means to them, what gratification they get out of it. It is necessary to distinguish here several levels of analysis, only two of which will be discussed briefly as examples for the others. One is the level of those reasons which are either conscious or could be made conscious. The best example is the list of reasons Mr. Rorty3 set up for people listening to commentators. It is quite true that detailed interviews can often lead to surprisingly new reasons of this kind. But it probably needs very gifted introspection to be a good respondent for such an interview. The best procedure, it seems, is to get a list of such reasons by interviewing a small number of people in considerable detail and then trying to find criteria by which less articulate people could be classified according to this list.
The use to be made of classifications of reasons on this level is probably two-fold: on the one hand it is interesting to know the distribution of these reasons in larger populations or in special groups. On the other hand, we might want to relate these reasons to other factors, for instance, effects: Do people who listen for different reasons take different things out of the program, like different parts of it, etc.?
5.) It is not quite certain, however, that a reason-analysis on this level will ever lead to very important results. The mystic desire of all modern psychology is to get at the »deep« stuff. By that is meant the use of such concepts as escape, repression, ego enhancement, compensation for inferiority feelings, etc. It is important to see the difference in the methodological approach compared with the previous paragraph. Those unconscious reasons can never be found by detailed interviews; it is necessary to assume them theoretically and then to prove or disprove their existence. Therefore, such interviews are so very difficult and even dangerous. The student really has to assume the general existence of such a mechanism and then to look for little details in the interview which point in its direction. Such interviews will mostly use the procedure of »Deutung«. Sometimes an inflection of the voice, a hesitancy to answer will be more important than the whole rational context of the interview. Only a student who has a well set-up theory will do a good job and only the most mature members of the staff should be using it. The main difficulty is that such results cannot be verified and, therefore, their whole value stands and falls with the value of the theory for our project as a whole. Empirical research here serves much more to exemplify the theory than to discover facts. Continuous discussion and clarification of such approaches will be necessary.
6.) A generalization of the reason-analysis is the personality approach. The idea here seems to be that if we know a lot about a respondent then we will understand how his radio activities fit into everything else he does. Therefore, the purpose of the detailed interview in this respect is to learn as much as possible about the respondent himself. That can be done either by tests or by other sets of comparable indices or it may be attempted by a looser description of each individual in his own right. An example of this approach is found in the relation of an individual’s radio habits with his general attitude toward music, his general attitude toward politics, toward education, etc. The main effort of the student will be to develop new and appropriate concepts to describe the different areas of personal activities. The results will most likely take the form of typological classifications: different types of personalities will be related with different types of radio attitude. Therefore, in this connection, the logic of such typological classification will be of special importance. That is also true for the next use of detailed interviews.
7.) It is probably an extension of the personality approach when we use life histories. Instead of using a personality test or a general description of the individual as we find it now, we try to understand him by following him through the course of his life. His musical or political development, his educational pursuits at different periods of his life are used as personal indices.
It is probable that in addition to this, life histories can be used to trace the influences of cultural and individual factors on his personality. In this way, we account for an individual’s radio habits not only in terms of his personality but also in terms of those factors which molded his personality, in those respects which are especially important for a radio study. While the logic of the personality approach has been studied extensively in the last decade, the biographical approach has only been touched incidentally. Therefore, the project will probably take the psychology of personality from the best sources and leave it at that, whereas in the life-history field we might be able to make a new contribution. There is actually only the book of Dollard on life-history available4 and the use which Blumer has made in movie studies5 and Gray–Monroe in the reading field.6 An organized effort to clarify the technique of using life histories should be made.
The most important task of our work in the field of detailed interviews will be to lift it to a more systematic level. So far, two points can be set forth, the first being of very great practical importance.
a) The principle of crucial groups. Just because detailed interviews are so laborious, it is necessary to use them so that they can be put [to] their maximum utility. Therefore, one should always try to interview people who have been pre-selected by preliminary information. If, for instance, we want to know why people like programs, we should interview people who like a certain program especially and those who exhibit a special dislike. If we want to analyze in detail the influence of radio on political opinion we should select people who actually have changed their opinions recently. An especially good opportunity might come up in our reading study as Miss Curtis has discovered.7 Alvin Johnson has studied the reading records of thousands of people at the Newark library and distinguished several types: those who read for escape, for specific improvement, for general cultural purposes, etc.8 Disregarding the value of just this distinction, we will, of course, interview people of distinct reading types if we want to know the relation between reading and listening; non-readers will, of course, have to be included. It is evident that such an approach will be much more economical than to interview just at random on the same topic.
It will be the task of all the staff members to find ways to select such crucial groups. If no information is available and a general approach indispensable, then at least accessory information should be gathered in order to distinguish different groups later on. If, for instance, we ask people whether they prefer straight news reporting or commented news over the air, we should try at the same time to get information on the recency of their political interests or on the independence of their judgment in other fields so that we may have a chance to later on pick out representatives of selected groups for detailed personal interviews.
b) In studying the attitude of people toward programs, it is useful to distinguish three kinds of studies. Those studies which are mostly concerned with the features of the programs; those which are especially concerned with the influences brought to bear upon people; those studies which are mostly concerned with the tendencies, needs, and desires of the listeners. The analysis of features and influences has advanced considerably in recent years and they are well taken care of in many parts of our project. What is still in bad shape is the analysis of tendencies, and evidently the entire use of detailed personal interviews centers around this problem. It will be useful to keep this distinction in mind in order to have the focus of those interviews well set. As a very good exercise for increasing one’s awareness of all the possibilities, and for improving the outline given above, a reading of Chapter 15 of Cantril–Allport’s Psychology of Radio9 is suggested. The ideas treated there are probably the most advanced material available in this field of analysis. By matching them with the approaches outlined above, the next steps for progress might be visualized.