Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
QUALITATIVE ATTITUDE ANALYSIS A TECHNIQUE FOR THE STUDY OF VERBAL BEHAVIOR RHODA METRAUX Formerly Technical Assistant, Committee on Food Habits Since early February ~94~, the Committee on Food Habits has been pre- l~arin:, a series of qualitative studies of consumers' attitudes toward changes ill the food situation and toward some of the programs initiated by the Gov- er~n~ent to meet the resulting problems. These studies have had as aims ~) to gain insight into current attitudes, ~) to aid in predicting future trends of reaction, and 3) to discover some of the appeals which could be used most ellectively as incentives to cooperative action. The present paper deals mainly with the technique of analysis and the premises on which this technique is leased. I "Words are part of action and they are equivalents to action." * That is to say, language communication is a formal of behavior which has its meaning within the context of a specific culture and of specific situations within the cultural framework, just as does any other. Studies of verbal behavior are valuable, not because one can predict that, having said something, a person will merely say something else, but rather because it is likely that, having said something, a person will do something. And the two, saying and doing, are related actions. Words here have in part an initiatory force, a directive power. For this reason it is essential to the success of any cooperative enter- prise that the participants "speak the same language." When, in a new situation, behavior patterns are modified as the result of new experiences, the meaning of words likewise is modified. Cooperative action on an adult level will then be possible only to the extent that the modifications are under- stood and accepted by all those involved. Changes in the production, distribution, and availability of food resulting from the war have brought about a new situation in the United States, espe- cially as far as the consumer is concerned. Although certain aspects are familiar to many individuals and groups, either through personal experience or through reports, nevertheless the present combination of circumstances has created what must be considered a new complex. It is to be expected that varied reactions to any single event will occur and that these will be variously interpreted, especially in the earliest stages of change, but, assuming a basic cultural unity, that differences of reaction will be more often (although not altogether) of degree than of kind. Understanding of both types of va- riation and of their relative importance is essential to the effective planning and smooth operation of any food program. One of the basic problems of * Malinowski, Bronislaw. Coral gardens and their magic. London, Allen & Unwin, 1935. Vol. 2, P. 9, et passim. 86
Qualitative ~ttitud e ~lna~ysis 87 leadership is that of making the situation intelligible, of aiding in the realign- ment of forces that lead to cooperation and efficiency. As in any emergency, it is necessary in the first place to work with the materials, the beliefs and concepts already existing, and the skills that exist or develop. Hence it is necessary to have a constant stream of information about how the people affected react, what older associations they call upon to meet the situation, where the positive strengths lie and how these are used, and where and when negative behavior limits the ability to cooperate and whether this is due to faulty understanding, incomplete knowledge, confusion or substitution of goals, etc. In this connection, studies of verbal behavior are particularly useful because they can be carried out on a comparatively large scale within narrow time limits and because they can give direct insights into associations and symbolism, that is, into the emotional and conceptual background of behavior. Studies of verbal behavior may be validated by comparisons with other types of behavior analyzed at the same level. Since no single event, but a series of interrelated events is involved in the development of the various food programs, studies of reaction must have a definite predictive value. Consequently, not opinions as such, but the under- lying attitudes must be the subject of study. Since the process of change is continuous, although now one aspect and now another may be highlighted, the method of study must be of a kind in which data can be assembled and analyzed rapidly and which allows a continuous comparison. Finally, since no single group but rather the total population is affected, the aim of the analysis should be to obtain a dynamic synthesis. II None of the current methods of opinion analysis quite fitted the purpose which the Committee had in mind. For example, a statistical approach seemed unsuitable to qualitative studies where the emphasis was not to be on, "How much?" or, "How many?" but rather on, "In what direction? In what re- lationships? With what implications?" Because of staff? limitations, it was necessary to depend largely on volunteer assistance in obtaining data from different areas of the country. This meant that the samples obtained might vary in size, but that a variety of skills could be called into action for obtain- ing data. A number of the collaborators available were highly skilled; the majority were students working under the direction of social science de- partments of colleges and universities; in addition, a number of other in- dividuals or whole groups could be asked to help from time to time.* A * Grateful acknowledgment is made to all those who gave so unsparingly of their time and energy in the collection of the data. We are especially indebted to the student squads who worked under the able leadership of Dr. Norma Bird of American Uni- versity, Washington, D. C., Mrs. Florence Booth of Wayne University, Dr. G. Gordon Brown of Temple University, Professor Gustav Carlson of the University of Cincinnati, Dr. Paul Cressy of Wheaton College, Dr. John Gillen of Duke University, Dr. Elton F. Guthrie of the University of Washington, Mrs. Jane Hanks, Dr. Weston La Barre of New Jersey College for Women, Mrs. Ethel McCormick, Miss Dorothy Johnson of the University of Michigan, Dr. Clyde Kluckhohn of Harvard University, Miss Helen Lockwood of Vassar College, Dr. Edgar A. Schuler of the State University of Louisiana,
88 Tl~e Problems of Changing Food Habits technique was therefore devised which did not require the use of a fixed population sample, which could make full use of the varied skills that were available, and which, to some extent at least, fulfilled the requirements of the aims described above. Briefly, the method of obtaining data is as follows. After formulating the problem, an open-ended question* or, more rarely, a simple statement is worked out which contains certain key words or phrases related to the gen- eral problem. Wherever possible these key words are ones which have al- ready been used by respondents in earlier studies or which are in current use to describe a program. Time limitations, unfortunately, seldom allow pre-testing. The question, together wild instructions for interviewing and for obtaining information about the respondents, is sent to the various inter- viewers who then go about getting answers from members of their respective communities. Occasionally they are asked to interview specific groups, i.e., grocers, persons waiting in line to register for ration cards, but as a rule the selection is not so limited, and investigators are asked to reach as varied a group as possible. Frown tinge to time students of different age groups and of varied socio-economic backgrounds, men or women in clubs, women in nutrition courses, and so on are asked to write brief statements in answer to the same question. When specially skilled assistance is available, data based on related questions and other types of material can be obtained. Such information is often of great value as a check for interview material. As far as it is possible, all interviews are recorded verbatim, since only in this way can an accurate picture of verbal behavior be obtained. Interviewers also are instructed to note the replies of those who, for one reason or another, refuse to answer the interview question, since such replies may give valuable indications of negative reactions. Since the time limit for obtaining the data is invariably short, the replies may be considered to represent the trends of atti- tudes prevalent in the sections of the country sampled at one period. In passing, it should be said that one of the great advantages of using, Dr. Afif I. Tannous, Miss Jane Leichsenring, Miss Mary Inman and Miss Lucy A. Studley of the University of Minnesota, Dr. Charles Vaughn of the University of Arizona, and Mrs. Pauline Park Wilson of the University of Alabama. For their cooperation and the contribution from their students and assistants, we are also greatly indebted to Dr. Theodora Abel of Letchworth Village. Thiells. lit Y.. Miss Marv Anderson of Washington. D. C.. Dr. Elsa ChaPin of Brooklyn, N .Y., ~, , , - _ , . Dr. Ethel Cornell of the University of the State of New York, Dr. Edwina Cowan of Wichita, Kansas, Dr. Mary Fisher, Dr. J. K. Folsom and Dr. J. L. Stone of Vassar College, Dr. Bertha M. Luckey of Cleveland, Ohio, Dr. Jean MacFarlane~and Dr. Ralph H. Gundlach of the University of California, Dr. Helen Peak of Randolph Macon Women's College, Dr. John G. Pilley of Wellesley College, Dr. Hortense Powdermaker of Queens College, and Dr. Sophie Robison of the New York School of Social Work. Grateful acknowledgment is also made to Mrs. Julia Katz, National Director of the i_ congress of Women's Auxiliaries, C. I. O. and to the Women's Auxiliaries in Harviell, Park City, Muncie, and Rachel who so generously contributed their help in obtaining data. We are indebted also for their cooperation to Mrs. Grace Porter of Aircraft Times and to the many members of the Aircraft Women's Clubs who contributed information. * See Appendix A for the types of questions used during the past year.
Qualitative Attitude Analysis 89 student interviewers lies in the fact that the answers given them appear to be rather free of some of the biases and inhibitions exhibited in more "official" contacts. On one occasion when some interviewers. inadvertently gave the impression of asking questions directly in behalf of the Government, the replies of the informants contained many phrases like, "You have to be care- ful what you say to the Government." The essays of school children, espe- cially those by children of upper grade school and lower high school age, also have a particular advantage because of the ways in which they reflect cur- rent adult opinion, in vocabulary as well as in formulation. It should perhaps be added that such essays, as well as all other statements, are written without preparation or previous discussion of the subject at the time of writing. When the data have been assembled, the preliminary analysis proceeds, community by community, or if it seems desirable, section by section within a community. For example, where the material from one community consists of essays by children of different ages, statements by members of a nutrition course, and a number of interviews, each group of replies will first be handled separately and then they will be integrated. To get a "feeling" of the community and to insure accuracy in the tabulation of vocabulary, it is important that all the material from one community be in the hands, or one might better say, in the mind of a single analyst. In a general way, each reply may be studied as a whole to judge whether, for instance, the informant is in favor of, or critical of, or against whatever is under discussion. However, the real analysis centers about the breakdown of the reply into its component parts into the responses. Each response is recorded separately. For instance, the following statement contains five responses; the total reply is in favor of enforced government rationing: "I think enforced rationing is the only ~way to do it because everybody needs salve meat and with voluntary ratio~ir/,g there are always solve who cheat. This ~way the arsq~' gets what it needs and we can share the rest equally.'' A simple check-off system of recording is used; every time a response occurs it is checked off on the list. Tout where a single response recurs in the same sense in a single reply, it is only recorded once. This requires a certain amount of patience and attention to detail since it is important to note and to follow, as far as possible, the precise vocabulary noting where alternate words are used with similar connotations-and, in addition, the relationship of one re- sponse to another within the whole statement. There is an obvious difference of pattern between two such statements as, ".... is netting short and so tarp chr~,lr] retire it,' ~nc1 "We cho',lel ration . · T T O ,^~ am _^,~,~ B~ to prevent a shortage." There is a more subtle but equally important difference between two such statements as, "~My boys are in the Navy and so I am willing to do with.- out so much .... ," and, "We should accept rationing of .... so that the boys in the Navy will have all they need." It is also important to prepare separate response lists for opposed types of answers. Thus, to anticipate a little, we will have in the end Patter;z ~4 which contains responses a, b, c, d, e, etc., and Patter;z B which contains responses ??*, A, O. p, etc. blare may be an overlapping of responses between two or more patterns; indeed, there usually is. Consequently, it is important that the lists be built up sepa
go The Problem of Changing Food Habits " 2 3 " 4 5 cc ~2 ` 3 4 c' rately. The breakdown, however, is far less difficult than it would appear to be, because there are, in fact, patterns of reaction which manifest them- selves in the use of certain types of expressions in similar sequences of recurrence. The use of a simple check-off system of recording does not preclude a simultaneous breakdown into age or sex groups or any other for which data have been obtained, since this can be taken care of by such devices as the use of special marks, different colored pencils, etc. The type of attitude data does not, however, lend itself very well to mechanical tabulation. When the preliminary analyses have been completed, the response lists from the various communities must be collated and drawn up into a total list. It is at this stage that differences in vocabulary or in emphasis in the various areas sampled should be noted, if they are significant. (This will already have been noted in some cases for sections of a single community.) Here again, in order that the range of attitudes and the relationships of the various elements to one another may be preserved, the tabulations must be worked out by a single analyst, preferably one who is entirely familiar with the replies as a whole, at least by having read through all of them. The material is now ready for the final steps of synthesis, diagrammatic repre- sentation, and interpretation. In the course of the breakdown of total replies into responses it will have been observed that the number of responses in the individual replies varies rather widely. That is to say, the key words in the question have evoked a varied number of associations in the different respondents. It is not to be expected that any one answer will contain every response in a pattern, but rather that the pattern or patterns will emerge from a larger number of cases. Schematically presented this means that in Pattern at: Reply ~ will contain responses a, b, c, d " " " b, d, e, f b, c b, b, Tic tic `` `` d, f' g, h, k d, e, f, h, and so on. In Pattern B: Reply ~ will contain responses m, n, o, r " " n, p, d (:~ ~ p, do) n, o, r, d (by ~and so on; {` tic `` m, Thus, it is clear that Pattern A is a, b, c, d, e, f, g, h, k. Or, since the emphasis on all points is not the same, Pattern At, correctly presented, is: a,5b,2c, 4d, 2e,3f,g, oh, k In the same way, Pattern B is: 2m, 3n, 20, 2p, Or, 3d (CAP) One response, d, appears in both patterns. The possible significance of this will be discussed below.
Qualitative Attitude Analysis 9I Actually, of course, the total range of responses will be much greater than the purely schematic arrangement a-k in Pattern A, and the differences of emphasis will be much larger and more obvious; for example, there might be 50b and 30f but still only one a response. In relation to b and i, therefore, a would be an aberrant response and would be knocked out at this point, as far as the composite statement is concerned (but see below). Reducing the re- sponses in this way until the core of greatest emphasis is reached the composite statement is arrived at. (The importance of a response made lo times will necessarily depend on whether loo or loon total replies are involved and a decision about what should be included must primarily be decided on this lapis.) This might be presented schematically for Pattern A: (glib, (2)C, (4)d, (2)e, (3)l, (2~ In building the composite statement, the arrangement of the responses is de- termined by their relationships to one another in the original replies. From this composite statement a diagram can be built up, the primary pur- pose of which is to show graphically the interrelationship of the responses and the variations in stress. The schematic arrangement of the elements in the diagram is determined by their relationship in the composite statement. The proportions of the various parts of the diagram in relation to one another are fixed by the number of times a response has been made. The accompanying statement and diagram taken from Beliefs about the Purpose of the Nutrition Programs in Fo?~ghkeepsie, N. Y. Area are given in illustration.* The graphic representation of the results need not follow any rigid formula. Indeed, it cannot. Before building a diagram it is necessary to go back to the problem as it was formulated previous to the collection of data, in order that the results may be finally organized and interpreted. It will be recalled that the response material almost invariably falls into two or more patterns; two or more composite statements will then appear. All of these are, of course, interrelated. How the statements will be presented graphically depends to some extent on the original problem. That is to say, while the relationships of the elements within a pattern or the relationships between several patterns are controlled only by. the material itself, there may be sev- eral ways in which these relationships can be expressed. It is here that prob- lems of interpretation first arise. A few examples will serve to clarify this statement. I. A similar response may appear in two opposed patterns (as d and d(~) in Patterns ~ and B above) for example, the statement, "Rationing is good because it wakes people up to the tear,' as opposed to, "I don't like rationing It is only done to wake people top to the war." Here we have an overlapping of responses which should be highlighted in the presentation since, however logical within the context of each pattern, the argument is one which has al- ready roused opposite reactions and therefore should be avoided as an appeal. This must be made clear to those who will make use of the results. a. An integral part of a program, which it is necessary for people to un- derstand if they are to cooperate fully in carrying out the program, may be * See Appendix B of this report.
92 The Problem of Changing Food Habits mentioned by no more than three or four out of 500 respondents. Ordinarily, this response would be knocked out in snaking the synthesis (see a in Pat- tern ~ above). In this case, however, it is important, because of the prob- lems involved, to see where it fits into the pattern and at the same time to show how it compares to other responses in importance. It should therefore be included in the graphic presentation. 3. The basic points of view expressed in two patterns may be so different that there is a special value in showing those points of contact that do exist and, more particularly, the areas of differentiation. This calls for a special form of presentation. Or, 4. The basic points of view about alternative programs, i.e., voluntary vs. enforced rationing, may be so similar that, once again, there may be a special value in overlaying, so to speak, the two patterns to show the basic similarities, especially the beliefs or values, that underlie both. Each of these illustrations involves a special problem of presentation. Others might be added, but it is already apparent that in every case cited, the ques- tions involved in the original problem as well as the results themselves affect the form of the diagram. Consequently, while the diagram is designed pri- marily to help those who are not familiar with the raw material to understand the results of the investigation, and while nothing can be shown in the dia- gram which is not inherent in the data, nevertheless its preparation involves a certain amount of preliminary interpretation. It is for this reason that the analyst must be entirely familiar with the data. ITT The larger problems of interpretation take one back to the original premise that language and action in a specific situation are linked forms of behavior. Both are moulded by the system of values, the beliefs, currently accepted. To discover what these beliefs are and where they are modified or even discarded over a period of time are major problems of interpretation, since a knowledge of values is essential to efficient long-term planning. Studies of verbal be- havior are useful because, in expressing their opinions about a specific event, people most succinctly refer to the underlying premises on which opinion and action are based. Such studies also bring to light emotional conflicts which may lead to apparent contradictions between expressed ideas and actions. For example, a respondent says, "I think rationing is a good idea because everyone gets his fair share; but there are always some people who cheat." One might expect the respondent to accept rationing with a good spirit on the basis of his answer; however, he has also expressed a fear which conflicts with his notion of equality. As long as this fear remains active, one may ex- pect him to take measures to protect himself to get what he considers to be his "fair share." The apparent contradiction between premise and (: observed) behavior is dissolved by an understanding of the fear. Why the fear arises at this point and why it is turned against some member of the community ("rich" people, defense workers, and grocers were among those commonly accused in the early stages of rationing) are further problems of interpreta- tion which a study of the total elements in the response patterns may resolve
Qualitative Attitude Analysis . 93 to some extent. A response of the kind described is important only for the individual if it is made by one person alone. Where such a response repre- sents the attitude of many persons in many communities, it has great signifi- cance. Composite statements and diagrams bring to light the existence of such premises and contradictions on a large scale. A. . . , . . , . , , , , ~ . . . ~ ne study of several re~arec~ sets of statements and diagrams snows now beliefs are affected by events; how older beliefs are called upon to make a new situation understandable; how such belief s are modified and where new beliefs replace them. The studies we have made have suggested, for instance, the existence of a strong belief in competitive democracy. In addition to the argument that rationing is good because it gives everyone a fair share or because it prevents some people from getting more than their share. an oc- casional argument against rigid rationing has been, "Why should the army get more than we do ?" More recent studies have suggested a modification of this belief in terms of need. Speaking of a growing shortage arid of the method of distributing the food in question, some respondents have said, "Give it to those who need it most." Thus it is clear that a greater flexibility has been attained which can be made use of in the development of future ~ ~ a. . ~ ~ ~ . ~ programs. There also have been indications of an increased awareness of the need to include the allies in the sharing process. However, faced by an actual rationing program, the awareness tends to drop off for a time. This belief appears not to be very strong, and consequently needs to be bolstered up by factual appeals, especially when personal deprivations must be faced. Other examples could be given. As the war progresses, new situations of many kinds will arise and changes of vital concern to the people of the United States will continue to take place, probably on an ever increasing scale. It is to the interest of all that the course of action taken in each case be decided upon in a democratic man- ner so that the Government and the ordinary citizen may cooperate on a [basis of mutual understanding. The use of the technique of analysis of verbal behavior described here is one means of providing a basis for the clarifica- tion of public attitudes. APPENDIX A Questions Used in Surveys Why is the Government laying so much stress on nutrition? What do you think about this business of registration? To win the war it is important to include housing, transportation and food in the Government's war program. What do you think about this statement? What do you think can be done about the meat shortage? Which foods do you think are most important for us to have to keep up our morale ? Why ? Which foods do you think are most important for us to have to keep up our spirits ? Why ?
94 The Problem of Changing Food Habits What is your most important food problem today besides the cost of food? What do you think can be done about it? What do you think is the most important civilian food problem today besides the cost of food ? What do you think about rationing ? What kind of cheese do you like to eat? I understand there is going to be very little cheese available for civilian use next year. What do you think ought to be done about the small amount that is available then ? Why ~ Why do you think cheese is important anyway ? APPENDIX B Nutrition Eat the Right Food GENERAL STRONG BODIES PHYSICALLY F I T _ RESISTANCE TO DISEASE AND EPIDEMICS 42 SOON I) MINI D I N STRONG BODY 34 GOOD MORALE 271 ENERGY EFFI CIENCY ALERT NESS 25 65 |HNALTH | VENOM] \ l I FREE DOC TORS 10 GOOD FIG HTERS 58 ICITIZENS GOOD DEfEN SE WORKERS 41 | \ ~\ STRONG CHILDREN 38 \ . \ GOOD RATION I NG 48 . ~ . COMBAT HIGHER COSTS OF LIVING 4 7 5 GROW RIGHT \ I \ FOOD CEl ~ MONEY FOR FOOD FOR DEFENSE / OTH ERS . ~ - win _ THE WAR 19 I . WIN Pl ACE FUTURE 15
