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.
FORCES BEHIND FOOD HABITS AND METHODS OF CHAN GE * KURT LEWIN I NTRODU CTIO~- The following report is a partial account of a study conducted by a field staff tt at the Child Welfare Research Station of the State University of Iowa. The objectives of the study were twofold; namely, ~) to investigate some aspects of why people eat what they eat, arid, if possible, 2) to study methods of changing these food habits. In studying the first question we have tried to combine approaches of cul- tural anthropology with quantitative methods of psychology. Two techniques were used. One is the Ba~relas testy a projective questionnaire,5 with which 2,300 school children were tested.§ The second method consists of interview- ing housewives. Five groups were studied; three representing economic sub- division (high, medium, and low income levels) of White American stock, and two subcultural groups, Czech and Negro. Several methods of changing food habits were compared experimentally. One experiment compares the effect of a method of group decision in one case with the effect of a lecture. Another experiment compares the effect of group decision with the effect of a request. The results of the interviews with housewives are presented first. I. WHY PEOPLE EAT WHAT THEY EAT A. Method: of Interz~iewi'*g and Analysis After a period of preliminary trials of various methods, the final data were collected during May and June, ~942. It should be kept in mind that the results describe the attitudes and habits of the people at that time (only sugar was rationed). The material was collected from the residents of a midwestern town with a population of about 60,000. Although surrounded by farmin, country, the town has a variety of industrial plants. It has employed a nutri- tionist for a number of years and has had a good nutrition program. * This study-was financed by a grant from the Committee on Food Habits. t The staff consisted of Alex Bavelas, Leon Festinger, Myrtle Hubbard, Patricia \Vood- ward, and Alvin Zander. Alex Bavelas acted as "discussion leader," Myrtle Hubbard as "expert" or "lecturer" in some of the experiments. One experiment was conducted by Ben Willerman. Alvin Zander, Beatrice Wright and Patricia Woodward did a major share of the analyzing and writing. "Indebtedness is expressed to Dr. Sybil Woodruff, Head of the Department of Home Economics of the State University of Iowa, and to Prof. Lynn Garwood and Dr. Alice B. Salter, of the Department of Sociology, Coe College, Cedar Rapids, Iowa. § A preliminary report 2 has been made but complete results are not yet available. 35
36 The Problems of C1tartying Food Habits The interviewers, trained and carefully supervised by a member- of the research staff, interviewed approximately an equal renumber of housewives in the three income groups and tl~e two subcultural groups. The income of tl~e low group did not exceed $~:500 per year; that of the middle group did not exceed Woo per year, and the high group was scattered, with the majority considerably above the upper limit of else middle group. The Czech women had been born in Europe and are comparable in economic level to the middle income group. None of the low income and Negro groups was on relief, but all were living on marginal incomes, although there is evidence Blat the Negro group is representative of a higher social level than the Sulfite group. Each interviewer talked alternately.with women from different groups so that no constant error due to the interviewer's personality or increasing routine would affect the results. A total of ~o7 housewives were interviewed, with approximately 20 in each of the five groups. Each interview lasted from 40 to 75 minutes. The women ire the high and middle income groups and in the Cycle group were chosen as "typical" of that group by other residents of flee city. The members of the low income group (White) and of the Negro group were selected by the director of the Community House serving the neighborhood in which most of them lived. In almost every case, the person who selected the interviewee explained to her that someone should be coming to see her. This considerably facilitated rapport. One of the outstanding difficulties in analyzing data about foods is to find categories which have meaning from the nutritionist's point of view and still are in line with the everyday terms in which the housewife thinks and acts. The thinking of the housewife is guided by quite a variety of aspects such as health, money, food for husband, breakfast food, food for Thanksgiving, ete. It seetns to be impossible to get a perfect classification system which takes in all these aspects at once. We have followed a line which, on the whole, tries to keep close to the thinking of the housewife.* B. Channel Theory . General consid equations. The question "why people eat what they eat," is rather complex, involving both cultural and psychological aspects (such as traditional foods and indi- vidual preferences caused by childhood experiences), as well as problems of *The following are the 25 categories of food used for groupie=, the answers to the questions: bread (bread, rolls, buns, biscuits; does not include toast); butter; casserole (includes foods frequently prepared in a casserole, as spaghetti and macaroni); caf3Seins (coffee, tea, coca cola); cereal (dry and cooked); cheese; desserts (pies, puddings, cus- tards, cakes, etc.); eggs; fish; flour; fowl; fruit; fruit juices; leftovers; meat (excluding fowl); milk (includes milk, buttermilk, cocoa, chocolate milk); potatoes; relishes (includes mustard ketchup. other seasoning. Pickles. spiced fruits. and vegetables): salads (fruit ~ _ 1 ~ ~ 1 ~ and vegetable salads); sandwiches; shortening; soup; sweets; toast; vegetables (lncluci- ing tomatoes but not potatoes). ~ A diagrammatic representation of the Clza~znel Theory appears in "Studies in topological and vector psychology III," edited by Kurt Lewin, to be published by the Child Welfare Research Station, State University of Iowa.
Food Habits and Methods of Change 37 transportation, availability of food in a particular area, and economic con- siderations. Therefore the first step in a scientific analysis is the treatment of the problem of where and how the psychological and the non-psycholo~,ical aspects intersect. This question can be answered, at least in part, by a "channel theory." Of paramount importance in this theory is the fact that once food is on the table, most of it is eaten by someone in the family. Therefore one would find the main answer to the question "why people eat what they eat," if one could answer the question, "how food coiner to the table and why." Food comes to the table through various channels. One is buying in a store. After the food has been bought, it may be stored in a locker to be taken out later, then to be cooked and brought to the table. Another channel is garden- ing. There are additional channels such as deliveries, buying food in the country, baking at home, and canning. a. Food Cozies step by step through a channel. The number of steps vary for different channels and for different foods within the singe channel. The time food can remain in one position varies. Food in the locker or food after canning may remain for considerable time ire the same position. Food may stay just a few hours or days in the pantry or in the icebox. To find out what food comes to the table, we have to know how many food channels exist for the particular family or group. To understand the changes after certain channels are blocked, we have to know what new channels open up or in which old channels traffic is increased. For instance, when preparin:,- meals at home becomes difficult, eating in restaurants may increase. Food does not move by its own impetus. Entering or not entering a chan- nel and moving from one section of a channel to another is effected by a "gate- keeper." For instance, in determining the food.that enters the channel "buying" we should know whether the husband, the wife, or the maid does the buying. If it is the housewife, then the psychology of the housewife should be studied, especially her attitudes and behavior in the buying situation. b. The forces governing channels. It is very important to realize that the psychological forces which influence the movement of the food may be dif- ferent for the different channels and for the various sections within the same channel. Each channel offers a certain amount of resistance to movement, and certain forces tend to prevent entrance into the channel. For example, if food is expensive, two forces of opposite direction act on the housewife. She is in a conflict. The force away from spending too much money keeps the food from going into that channel. A second force corresponding to the attractiveness of the food tends to bring it into the channel. Let us assume that the housewife decides to buy an expensive piece of neat: the food passes the gate. Now the housewife will be very eager not to waste it. The forces formerly opposing each other will now both point ire the same direction: the high price that tended to keep the expensive food out is now the reason why the housewife makes sure that through- all the difficulties the meat gets safely to the table and is eaten.
~,8 The Problem of Changing Food Habits ao up cat ~ _ ct us ~ . .= ct Q Us Us ~O ~ in be * _ V: U) o ,= o P; 5: O O o 1 i_ P; me" :Z ~ . V ,c; U? C) A: Cal so 3 ~._ O ~00 NO \O ~.= ~ a.~= x 0 ~ 0 ~ ~ O ~ oo O I I 1 1 1 cat v) .0 00 ~ cat ~ to ~ on 00 ~o0 ~cat ~ O Cx ~0 ~u ~- 0 c~ C ~O ~00 ~04 "S ~0 r. V ~ - 1 . ~ _ i_ . CN G~ O O ~O O 0 1` ~1` _ ~ O O O 0 00 00 ~D ~ Cx C~ CO O O O~ ~4 ~ ~ ~0 o0 0\ 00 ~4 O S~ 00 C~ o o ,~ Ct ~i 1 ~ .m .bC l ~ ~ C~ - D C~ U. - '_~ ~; C) UO V] -4 ~C tn C) bC CC _ CJ *
l The use of caribous channels. Food Habits and; Methods of Change 39 vow v an, em, van vv~ - , verve - ~vv 'l'he question used in the interview to obtain information on the different channels used by the housewife was expressed in this form: "People obtain food in many different' ways. They may buy it in the store, or farm it, or can it. We are interested in finding out what these ways are. In what way does the bread you use come to your table?" If several channels, such as buying and baking, were named, the interviewee was asked to estimate the proportion of the amount coming through each. Information was obtained for each of the following categories of food: ~) dairy products (eggs, milk, cheese), 2) fish and meat, 3) vegetables, 4) fruits, 5) desserts (cakes, pies), 6) bread (all forms). If different foods in the same category, such as eggs and milk, came through different channels, this was recorded also. The results show that in the five groups studied, each of the foods, except desserts, is obtained through the buying channel considerably more frequently than through any other channel, as shown in Table I. However, there is a difference between groups in shopping habits (Table 2~. On the whole there is a tendency to buy less than three times a week. The high income group is the only one which tends to buy more frequently than three times a week. The well-to-do housewife may not be as concerned with economizing on energy and time and is more likely to have a maid. T A B L E 2 PER CENT OF PA MILIES B UYING AT V ARYING INTERVALS Total High Middle Love Czech Intervals N-~o6 N-~3 N-I8 * N-~T N-23 More frequently (at least 3 times per week) ......... 25 % 405S I75S 245S 1870 Less frequently ( less than 3 ~ times per week) 40 27 33 Both more frequently and less fre quently 35 35 50 * Information on this question was not obtained from one person. Negro N-2~ 29% 34 57 48 43 27 24 For all groups together, about a third of the vegetables and fruits are canned at home. There seems to be no relation between income levels and percentage of families who can. However, there is a cultural difference. All of the Czech families do some canning; the difference between them and every other group is significant with the exception of the low income group (Table 3~. This, T. A B L E 3 PER CENT OF FAMINES USING DIFFERENT C HANNELS Total High Middle Low Czech N-~o7 N-23 Nag N-2I N-23 Home canning 83 74 79 go IOO Have gardens ............. 5 ~7 47 57 87 Negro N-2T 7I 52 along with evidence to be presented later, suggests that the Czech group is more self-sufficient than any of the other groups. We find the amount of food canned to be greater in the two lowest income groups and the Czech
4o The Problem of Changing Food Habits TABLE 4 MEAN PER CENT OF FOOD CANNED AT HOME Vegetables ................ Fruits ......... Two highest Two lowest income income groups groups and Czechs do ......... IS .................... I4 4s than in the two highest income groups (Table 4~. The most frequently canned vegetable for all the groups is tomatoes. The most frequently canned fruit is peaches. There are also differences in the type of food canned by the va- rious groups: jams and jellies are canned by a significantly greater percentage of families from the middle and high income groups than from other groups. Noodles were canned only by Czechs; one woman showed the interviewer hundreds of quarts of noodles. These results may be interpreted to mean that the motive for canning varies among the income groups. The lower income groups seem to be more guided by financial considerations and can essential foods whereas the high income groups can for taste and possibly status. Table 3 shows that the two lowest income groups and the Czechs do sig- nificantly more gardening than the other two groups. Lockers are used by almost 5070 of the high income group, but by no fami- lies in any of the other groups. The above discussion on food channels shows that to some extent financial circumstances and cultural values do influence the extent to which various food channels are used and the uses to which they are put. Thus, the lower income groups are able to effect savings by canning more of the essential foods they eat and by having more food gardens. The higher income groups are able to maintain lockers and have milk deliveries. The Czech group, re- sourceful and strongly motivated toward self-sufficiency, does the most can- ning and gardening. 3. Who controls the channel. It is important to know what members of the family control the various channels, as any changes will have to be effected through those persons. In all our groups the wife definitely controls all the channels except that of gar- dening, and even there the husband seldom controls this channel alone. Children are never mentioned as controlling any of the channels, although they undoubtedly influence the decisions indirectly through their rejection of food put before them. C. The Psychology of the Gatekeeper To understand and influence food habits we have to know in addition to the objective food channels and objective availability, the psychological factors influencing the person who controls the channels. The psychology of the gatekeeper includes a great variety of factors which we do not intend to cover fully. The factors might be classified under two headings, one pertaining to the cognitive structure, i.e., terms in which people think and speak about food; and the other pertaining to their motivation, e.g., the system of values behind their choice of food.
Food Habits and Methods of Change 4I I. The cognitive structure. The cognitive structure deals with what is considered "food," "food for us," or food for other naen~bers of the family, with meal patterns, and with the significance of the eating situation. a. Food outside and within consideration. Physical availability is not the only factor which determines availability of food to the individual. One of the determining factors is "cultural availability." There are many edible materials which people never even consider for use because they do not think of them as food for themselves. If we consider as food all that which some human beings actually eat and like to eat, then live grasshoppers would have to be included in the category of food. If, however, we ask what people in the United States consider as food, live grasshoppers would be excluded. In other words, the psychological area of food in our culture is only a small part of the objectively edible food, and could be represented diagrammatically as a small restricted region within the total region of all objectively edible food. In some parts of our country peanuts or cheese are considered food for animals but not for human beings. A farm girl in Iowa refused to eat cottage cheese because it is something for the pigs. Even within the area of food in our culture, the boundary between food for human beings and food for animals varies. Even the food that is recognized as that for human beings still may not be accepted as food for one's own family. For example, kidneys or certain viscera are considered by some as food only for poor people, or champagne a drink for the very rich. In other words, only a certain part of the area recognized as "food for human beings" is recognized as "food for us." To find out what is considered "food for us" by different groups is one of the first objectives of studying food habits. b. Food for husbands and: children. Within the area of "food for us" one might distinguish "food for the husband" and "food for children" as special subareas. The fact that the housewife controls the channels does not mean that she is not influenced by the preferences of the husband, or what she thinks is good for him and the children. Their indirect influence is shown in answer to the questions: ~ . . . I. What things do you cook when your husband is home that you do not cook when he is away? 2. What things do you cook specially for your children that you would not cook otherwise? The most typical husband's food (Table 5) is meat, which is mentioned by 39 per cent of the families, a significantly * greater per cent than those ~ The tests of significance used are in all cases either critical ratio between percentages, chi-square tests, or t tests. The critical ratio between percentages and the chi-square tests were used in instances where the numbers to be compared were simple frequencies, each individual appearing only once. In cases where one individual could appear many times, variability between individuals was calculated and a t test applied to the difference be- tween mean number of times a given thing was mentioned.
42 The Problem of Changing Food Habits mentioning vegetables. Meat ranks first as a husband's food for all the sub- groups except the Negro group where it ranks third, and vegetables and des- serts precede it. On the other hand, the most typical children's food is vege- ta0ies, mentioned by one-th~r~ ot the tamales having children. Vegetables ~ ~ ~ cat - rank first as a children's food for all the groups except the Negro group where it ranks second and desserts come first. Potatoes are served more frequently as a special dish for the husband than for the children. TABLE 5 RANK ORDER OF THE MOST :FREQUENTLY MENTIONED FOODS SERVED FOR HUSBAND AND CHILDREN PER CENT OF GROUP Husband Total group High Middle Low Czech Negro N-IO7 N-~3 N-19 N-OI N-~3 N-~T Meats 39 Meats 43 Meats 48 Meats 48 Meats 39 Vegetables 38 Vegetables HI Vegetables 3 ~Potatoes 3 ~Potatoes 33 Soups O6 Desserts 33 Desserts ~1 Desserts . I3 Desserts ~6 Desserts ~9 Casserole ..I3 Meats .I9 Potatoes 17 Cereals 9 Vegetables ~6 Vegetables T9 Bakery 9 Potatoes I4 Children Total group High Middle Low Czech Negro N-88 N-I6 N-I9 N-2I N-II N-2I Vegetables . . 32 Vegetables . . 3 I Vegetables . . 26 Vegetables . . 24 Vegetables . . 45 Desserts . . . 29 Desserts 23 Meats 3 I Desserts ~ I Desserts ~4 Desserts ~ ~Vegetables ~1 Meats I6 Desserts I9 Salads . 2I Casserole 19 Casserole I7 Meats fig SOUPS II Cereals I3 Meats I6 SOUPS I4 SOUPS 19 Potatoes 8 This indirect control by other members of the family is one of the many aspects of the psychology of the gatekeeper. c. `'Meal patterns." Other aspects of the cognitive structure of food are the difference between breakfast food, food for lunch, and for dinner; the distinction between main dish and dessert; the concept of balanced meal and of "leftover." The housewife was asked: "In what terms do you think of a meal: what goes into a breakfast? lunch? dinner?" As an answer she might say, "At dinner we have some kind of meat, a vegetable, salad, and a dessert," or "For lunch we either warm up leftovers, or try to have some other hot dish, as soup, with sandwiches and milk." Because of the limitations of space, the lists of foods mentioned by each group will not be given, but some outstanding differences between the groups will be described. Cereal, caffeine (coffee, tea), eggs, and bread or toast are the most generally accepted breakfast foods by all the groups. Fruits are mentioned by three-fourths of the high and middle groups, but by only one-fourth of the Czech, Negro, and low groups. As lunch foods, fruits and milk are mentioned more frequently by the high group, and soups more often by the low group. Salads, sandwiches, and fruits are much more characteristic of the high and middle income groups than of the others. Leftovers are used by more than half the Czech group, by about one-third of the high, low, and Negro groups, and one-sixth of the middle group. Lunch is apparently a "pick-up" meal more than either of the other meals. Whereas approximately 75~ of the high and middle groups claimed to plan
Food Habits and; Methods of Change 43 their lunches, only s~ to go of the other three groups did. The others said they ate whatever happened to be in the house. Meat, vegetables, potatoes, and dessert are commonly accepted by all groups as foods for dinner. Salads are mentioned much more frequently by the two highest income groups (6g70 as compared with ~940) while bread is listed less often, and butter not at all; ~5/ of the three lower groups name butter, and 3370 name bread. It is likely that bread and butter are considered a real part of the dinner in these three groups, and only accessories by the two higher income groups. One outstanding dimension of a culture is the degree of homogeneity of the individuals within it. A quantitative measure was developed which de- scribes the degree of homogeneity in respect to food habits. Obviously, the more divergent the responses to a given question, the more heterogeneous the culture since more leeway for individual preference is permitted. When a culture is exceedingly strong in its dictates individual differences will be minimized. An index which ranges from AN to Coo (N being the number of people in the group) was thus devised to reflect these relationships.* TABLE WHOMOCENEITY INDICES FOR BREAKFAST, LUNCH AND DINNER Foods High Middle Low Czech Negro N-23 N-I9 N-2I N-23 N-2I Breakfast ~5 ~4 ~7 no no Lunch 24 ~ o 7 9 6 Dinner ~7 fig ~8 ~5 ~5 Table 6 shows the homogeneity indices (H.I.) for the three meals. The larger the H.I., the greater the similarity among the foods mentioned by the members of one group. This in turn can be interpreted as meaning that the culture defines the "meal pattern" more precisely than in those cases when the H.I.'s are smaller. There is considerably less agreement about the pat- tern for lunch than for either breakfast or dinner in all except the high group. This indicates that the families of the high group agree more than do the families of the other two groups as to what foods characterize lunch. In regard to dinner, there is considerably greater agreement among the members of the middle group than among the others. The degree of homogeneity of meal patterns may be of particular im portance In regard to changes In food habits. There are indications that regardless of working hours, people eat according to the clock.4 A worker waking at noon to go to work will eat a lunch meal rather than a breakfast and nutritional elements which he ordinarily obtained through breakfast foods may be deficient in his diet. Moreover, since the results indicate that the lunches of all but the high group are the least "structured" of the three . . . . . ~ . ~ *The homogeneity index was calculated in the following way: the mean number of categories mentioned per individual was divided by the total number of different cate- gories mentioned by the group. Thus, the highest homogeneity index possible would be one. This would occur if all the individuals mentioned the same categories. The lower the homogeneity index, the less agreement there is among the individuals making up the group.
44 The Problem' of Changing Food Habits. meals, it would follow that other foods might be fitted most easily into the lunch pattern. d. The meaning of the eating situation. One important Loire is the feeling of group belongingness created by eating in the company of others. At a ban- quet, eating means something very different from eating after a long period .of starvation, and may be classified as a social function rather than as a means of survival. On the whole, eating is usually a more complicated function than just taking nourishment. The psychological meaning of eating is closely related to group situations. Eating with fellow-workers in a factory is something different from eating at the family table or eating in a restaurant. The "eating group" influences greatly the eating conduct and the eating ideology of the individual One can say that every eating group has a specific eating culture. a. Motivation. We will discuss the various factors in motivation under three major head- ings; namely, a) values (motives, ideologies) behind food selection, b) food needs, and c) obstacles to be overcome. a. Ualues behind food selection. There is more than one value which acts as a frame of reference for the individual choosing foods. These values have not always the same weight for the individual; they may change, as during wartime, and in addition may be different ire the restaurant and at home. The earlier, exploratory phase of the study indicated that at least four frames of reference are used in evaluating foods; namely, expense, health, taste, and status. It is important to know the relative strengths of these dif- ferent frames of reference for various groups of people and also how they vary for different foods. In regard to the system of values, three questions may be asked: ~ ~ What are the values for this group? 2) What is the relative weight of each value? 3) How are specific foods linked with certain values? I. What are the values? The usual informant is often unable to define the values which govern him because they are not thought of explicitly, but are part of a non-verbal implicit system of reference. Therefore we used two indirect methods for uncovering the value scales of the interviewee. The first consisted of watching carefully for all offhand comments to one of the fol- lowing frames of reference: a. Money, e.g. "Our family loves oranges, but we have stopped buying them. They are too expensive." b. Health, e.g., "My children have to have a quart of milk a day for their teeth." c. Taste, e.g., "I don't serve desserts at lunch time. We're not very fond of desserts." d. Status, e.g., "We have our meats sent from Chicago." Health considerations might be subdivided into the general idea of "good for you" or the more specific ideas of "good for teeth" or "vitamins." The
Food Habits and Methods of Change 45 values might lee considered relative to the person himself or to other members of the family. The interviewer noted such side remarks with a code letter in her running account of the interview, and wherever possible recorded the exact statement. This method has recognized limitations in that an interviewee may take a value so for granted that she may not mention it or may be unable to men- tion it in the particular interviewing situation. For instance, the status factors are probably much stronger than they appear to be. However, the results obtained by this method show clear differences between the subgroups and present a coherent picture. 2. Relative weight of dillerent frames of reference. There are significant differences in the frequency with which various frames of reference are mentioned both between the groups and within each group (Table 73. TABLE 7 AVERAGE FREQUENCY OF High N-23 1.30 2.61 I -34 .o8 I; rams of reference Total N-~o7 Money ~.80 Health 2.08 Taste .90 Status .o4 MENTION OF VARIOUS Middle N-I9 357 2.57 .63 o Low N-2I 3.81 1.38 .76 o FRAMES OF REFERENCE Czech Negro N-23 N-2 ~ 2.2 I 3.42 2.61 ]~19 ·78 ·95 0 ·14 Money and health came up significantly more often than other values for the group as a whole and for each subgroup except the high income group, whose taste plays an important role. Morley is mentioned less frequently by the high group than by the middle, low, Negro (significant below the 57o level), and the Czech groups (significant at the 87o level). It is more fre- quently mentioned by the middle than by the Czech group (significant at the ~70 level) despite the fact that they are at approximately the same income level. Health is mentioned significantly more often by the high and the Czech groups than by the low and Negro groups; significantly more frequently by the middle group than by the Negro group; there is no significant difference between the middle and the low group. Within the groups the following differences can be observed. In the high income group, health is the predominant value, with money and taste at a lower, approximately equal level. In the middle group money is the pre- dominant frame, with health considerably lower, and taste a great deal lower. This is also true of the low income group except that the differential be- tween money and health is even greater, money being by far the most im- portant consideration. This picture is in general the same for the Negro group. The Czech group falls between the high and middle groups, in that their mention of money and health are approximately equal, with taste a great deal lower. 3. VaYues and specific foods. In order to know which food will be chosen one has to know, in addition to the general value systems and the relative weight of each frame of reference, exactly where each of the foods in question
46 The Problem of Changing Food Habits stands on each of the value scales. The following questions (asked at the end oWthe interviews were designed to bring out the relationship between value frames and certain foods: a. What dishes would you be sure to serve if you were very short on money? b. What dishes would you be sure to serve if your only consideration was health ~ c. What dishes would you be sure to serve if.your only consideration was taste ~ d. What dishes would you be sure to serve if your only consideration was to put on a "fuss" and have a company dinner) e. What foods would you be sure to serve if your only consideration was to fill up some people. Table 8 shows the results for the total group in regard to their mention of meat dishes. Figure ~ gives the profiles for three specific meats. Fowl ...................... Fish ....................... Steaks and Chops.......... Roasts .................... Casseroles and Stews....... Glandular and Chopped..... Meat (general) ......... Meat (miscellaneous) .... TABLE WPER CENT OF TOTAI~ GROUP ( N-107) MENTIONING VARIOUS MEATS IN ANSWER TO FIVE FRAMES OF RETERENCE Meats Short on money Fill up Health Taste Fuss 6 2 I4 40 0 5 7 I 7 7 I8 6 II 5 8 23 5 2 2 2 I2 II 5 2 I4 I2 7 6 7 IO II 8 He a: ~ `,~ ~ 50Yo LL ~ ct LL .~ ~ J cow z ~ ar O z ~ Ill,! z ~ z cY - o cow 6 IS 6 9 SHORT ON FILL UP HEALTH TASTE FUSS MONEY 25 To ~ o onto . ~D ~ ~ TIC = FOWL ~ ~-~v = GLANDULAR & CHOPPED FIGURE I. Position of three kinds of meat along the five frames of reference for total group.
Food Habits and Methods of Change 47 Fowl is almost never mentioned as a dish to have when short on money, or as a most healthful, or most filling food, but is mentioned by 40%0 of the group as a dish to have for a company dinner. This can be interpreted as meaning that in May and June, ~94~ the housewife was not likely to buy fowl if she considered only money or health, but was likely to do so when buying for a company dinner. At the present time these considerations seem to have changed as a result of fowls not being rationed. The position of the various foods on the taste scale was investigated, in addition, by asking the housewife, "What dishes are your family especially fond of ?" This form was chosen because foods that are considered favorite by the entire family are more apt to reflect the effect of the general culture than the foods favored by an individual. The average number of favorite dishes per family was approximately six in each group. As Table 9 shows, meats, desserts, and vegetables are the most TABLE 9-RANK ORDER OF THE MOST FREQUENTLY MENTIONED FAMILY FAVORITE MOODS PER CENT OF GROUP High N-23 Meats 87 Desserts . . .48 Vegetables . . 3 5 Salads 30 Fish 22 Casserole . .2 ~ Middle N-I' Meats 79 Desserts . . .79 Vegetables. .63 Casserole . .53 :Bread 42 Fruits 3~ LOW N-21 Vegetables. . 76 Meats 5 7 Desserts . . . 5 7 Casserole . .24 Czech N-23 bleats 87 Bread 78 Vegetables . . 6 s Desserts . . .35 Soups so Casserole . .26 Fruits 22 COMPARISON OF DIFFERENT INCOME LEVELS High Meats 87 Vegetables Desserts Casserole ...... 48 ....... HI Middle 79 63 79 53 Negro N-2I Vegetables . . 9 5 Meats 67 Desserts . . .67 Bread 48 Casserole . .43 Fruits 24 Low 57 76 57 57 frequent favorites in all groups except the Czechs who name bread signifi- cantly more often than desserts. That this category, bread, is so high for the Czechs is probably due to their large consumption of kolatches, a Czech dish made of dough similar to bread and stuffed with meat or fruit. Meat tends to be less mentioned as a favorite dish with decreasing income level. Vegetable dishes show the opposite trend and are mentioned signifi- cantly more often by the low and Negro group than by the high income group. This may be interpreted as supporting the hypothesis that people like what they eat rather than eat what they like. Our -data do not give support to the widely prevalent idea that favorites are generally those foods which are difficult to obtain. The housewife was asked, "What foods do you think are essential to a daily diet?" This question was introduced to obtain an over-all reaction of the housewife to the importance of foods within the diet. The degree to which a food is considered essential might be of particular importance in planning changes of food habits.
48 The Problems of Caging Food Habits TABLE 10~RANK ORDER OF THE MOST FREQUENTLY MENTIONED ESSENTIAL FOODS - PER CENT OF GROUP High N-23 Vegetables. .96 Milk 96 Meats 9 ~ Fruits 74 Eggs 70 Potatoes . . . 43 Butter .... 3 o Bread 22 Middle N-~g Vegetables. loo Milk 89 Fruits 89 Meats 84 Eggs 63 Potatoes . . . s 3 Bread 37 Cereals .... Butter .... 2 I Fish 2 I Low N-2~' Vegetables. . go Milk s 7 Potatoes . . . 5 7 Bread 48 Meats 43 Butter .... 33 TO tggS 29 Fruits 29 Cereals .... 24 Czech N-23 Vegetables. . 83 Milk 83 Meats 83 Butter .... 52 Fruits s 2 Bread 52 - tggs · 3 9 Potatoes . . . 26 COMPARISON- OF DIFFERENT INCOME LEVELS Vegetables Milk ..... Meat ..... Fruit .... Eggs .... Bread .... Potatoes . . Negro N-2~ Vegetables. . go Milk Meats 71 Bread 5 2 Fruits 48 Eggs 48 Potatoes . . .48 Butter .... 3 3 Cereals .... ~9 High Middle Low 96 IOO 90 96 8g 57 9 I 84 43 74 89 29 70 63 29 22 37 48 43 53 57 As shown in Table IO, vegetables and milk are the most frequently men- tioned essential foods in all groups. Bread is considered essential by sig- nificantly more families from the low income group, the Czechs, and the Negroes, than from the high income group. Fruits are regarded as essential by four-fifths of the high and middle groups and by less than half of the three other groups; eggs by two-thirds of the high and middle groups and by about one-third of the others. b. Food reeds. It is important to recognize that the relative weight of the various frames of reference changes from day to day in line with the changing needs. These needs might change because of satiation, of variation in the situation, or because of cultural forces toward diet variations. It is in line with the basic phenomena of all needs that continued consump- tion of the same type of food leads to a decrease in the attractiveness of that particular food. This is a powerful force toward daily and seasonal cycles in food choice. It affects different foods in different degrees; for instance, it is smaller for bread than for meat. Willerman's study 6 indicates, however, that even in the case of bread, which is something like a "background" food, the desire for variety can be noticeable. The general level of food satisfaction, too, affects the attractiveness of food and changes the relative weight of the various value scales. If less food is within reach of a person the relative weight of the taste scale tends to diminish in favor of the "essential" aspects of food. If the food basket is pretty well filled the housewife can afford to be more discriminating in her choices than when it is empty. The situational factors are fairly obvious: when the housewife is short on Morley at the end of the month or when she is preparing a meal for guests, the corresponding frames of reference will increase in weight.
Food Habits and Methods of Change 49 The continued advocation of a "rich and varied diet" during the last decade has strengthened cultural forces toward day to day variations in foods. c. Obstacles to be ozrerco~ne. The interview did not approach the problem of obstacles along the various channels in a specific way, although these prob- lems must be taken into account in planning changes of food habits. Canned foods, for instance, are frequently preferred because of the little time neces- sary for preparation. The extent to which such obstacles as difficulty in transportation, lack of domestic help, time necessary for preparing and cook- ing influence the choice of the gatekeeper depends on his particular cir- cun~stances. 3. Conf ice. a. Buying as a decision situation. We have discussed a number of forces which act toward or away from choosing a given food. Their simultaneous presence in the actual choice situation creates conflict. In general a conflict situation arises when there is, on the one hand, a drive to engage in a certain activity (as buying food) and on the other hand, a force opposing that activity. An increase in prices, acting as a resistance to buying the foods which people have grown accustomed to, enhances the conflict in the food area for all groups. Families of low income are likely to experience more conflict in buying food than those of high income since their freedom in buying the foods they want is restricted by their limited ' finances. Members from the middle income group, however, may experience greater conflict than those from the low income group in so far as they are psycho- logically a marginal group. They strive to achieve the social status of the financially more able and at the same time fear dropping back to the level of poor people. The degree to which a proposed charge' of food habits happens to touch a food area of high or low conflict is one of the factors determining the degree of emotionality wish' which people will react. b. Conflict in a situation of rising food prices. I. Situation at time of study. At the time of this study in the Iowa town, sugar rationing was just beginning; there were very few shortages in other commodities. However, prices of foodstuffs had gone up without a comparable rise in income and people were especially conscious of the rising cost of food. Three questions concerning food retrenchment were asked: "if Which foods are you already cutting because of the increase in the price of food? ~) If prices continue to rise, which foods Night you cut? 3) Even if prices continue to rise, which foods are you particularly anxious not to cut?" As might be expected (Table ~ ~ ~ the high income group has cut fewer foods than any other group. The next in line are the Czechs. Both groups have cut significantly less than the middle, low, and Negro groups. The middle group has cut almost the same number as the low group, both rather heavily, namely, almost ~, foods per person. 4
So The Problem of Changing Food Habits TABLE 11 AVERAGE NUMBER OF FOODS NAMED IN ANSWER TO CONFLICT QUESTIONS Total High Middle Low Czech Negro N-IO4 N-23 N-18 N-~I N-~3 N-I9 Already Cut ~.73 .96 ~ 33 ~ 38 ~7 2.05 May Cut .7g .57 ~ ~·38 i.52 ·32 Don't Want to Cut 40 ~.3g 2.94 ~.24 1.87 2~74 ..... - - ·57 2.3g 2.94 1.52 1.87 In regard to foods they might cut the Czechs list more than any other group. At the same time they list fewer foods which they do not want to cut than any other group. Their willingness to get along on fewer foods may be partially due to their past experience. Their attitudes were expressed suc- cinctly in the comments of several of the housewives: "In the old country we were used to getting along on much less. It takes more time to fix it, but we can do it." "I'd make everything go farther. I always had to do that in the old country stretch things as far as they could go and do without meat except once a week." "In the last war we cut down on sugar, mixed butter with oleo, and stretched meat by using gravies and soups, and we could do it again." The low income groups can be assumed to be using a smaller variety of foods and consequently would realistically mention fewer foods which they might cut. The high group, on the other hand, in spite of their greater variety of foods, does not expect to cut much, probably because of their wider margin of financial security. 2. General intensity of confict. The following scale of conflict ratings was used in determining the total conflict rating for each individual. The term"food" refers to one of the 25 food categories used in classifying foods throughout this study.* I. Given to foods mentioned in answer to only one of the three questions. 2. Given to foods mentioned in answer to both questions ~ and a. 3. Given to foods mentioned in answer to both questions 2 and 3. 4. Given to foods mentioned in answer to both questions ~ and 3. The reasoning behind this arbitrarily established rating "scale" was that there is some conflict associated with a given food if it is mentioned in answer to any one of the questions, but that the conflict shows a progressive increase ~) if the food has already been cut and may be cut still further, ~) if the food might be cut but is one which the individual does not want to cut, and still more 3) if the food is one which has already been cut but is one which the individual does not want to cut. There may be some differences in the conflict expressed by an answer to any one of the three questions (rating of If, but such a distinction would involve a finer scale than was thought necessary for the purpose of this analysis. The maximum conflict rating that could be obtained by any one individual would be loo, that is, all 25 foods with a conflict rating of 4. Such a score is practically unattainable, however, since no individual has cut every type of food. The scores of the group studied ranged from 0 to ~3 and are shown in Table 12. * See footnote on page 36.
Food Habits and Methods of Change TABLE 12 AVERAGE CONFLICT RATINGS FOR THE FIVE GROUPS High Middle Low Czech N-~3 N-I8 * N-~I N-23 Average Conflict Rating a.~- 7.ad ~.6 ~Aft ~ v _ * One person did not answer these questions. ~¢ Two persons did not answer these questions. 51 Negro N-I9 ~ 574 The average conflict rating (Table ~) for the middle group is significantly higher than that for the high group and for the Czech group. The high in- come group, being the most secure financially, would be expected to feel the effects of the rise in prices the least and to have made the least changes be- cause of them. The middle group feels the rise in prices most for several reasons: ~) they are in a marginal position; the future holds more of a threat for the middle class; in the effort to resist lowering their social status they might economize first in those areas which are socially least prominent, such as food, thus keeping up appearances; 2) they are accustomed to de- pending less upon home canning and other economizing measures than are the lower income groups; 3) their dominant frame of reference is money (see Table 7) which is mentioned spontaneously by them almost as frequently as by the lower income groups. At the same time they emphasize health al- most as much as the high group and much more than the lower group. There- fore, a rise in prices can more easily bring about a rise of conflict for this group than for the others. 3. Confict ill regard to special foods. The relationship of the conflict rating to specific food categories is shown in Table ~3. TABLE 13 CONFLICT RATING OF FOODS FOR DIFFERENT GROUPS Total group High Middle Low Czech Negro _\ ~_ _> {_ r_ ~ G ~O G.= ~ G._ ~G. ~ G.= ~G.= Meats I.O9 Vegetables .87 ATe&retables 1.44 Meats .95 Meats 1.57 Meats Loo Vegetables .75 Milk .70 Meats :.~8 Potatoes .76 Vegetables .48 Milk .84 Milk . 6z Meats 65 Butter 94 Butter 67 Milk 43 Eggs 68 Butter 56 Fruits 43 Fruits 94 Fruits 62 Sweets 43 Butter 63 Fruits 54 Eggs 39 Milk 89 Vegetables .57 Fruits .... 39 Sweets 53 Eggs 37 Butter 30 Dessert T 39 Eggs 33 Butter .... 35 Vegetables .47 Sweets * .3z Potatoes .33 Milk .33 Bread ... .35 Desserts . Potatoes .30 Bread :: .33 COLIPARISO~ OF DIFFERENT INCOME LLVEI S Milk .... Meat .... Butter . . . Fruits . . . Potatoes High Middle Low 7o .89 .33 .65 T.28 .95 so .94 ·67 ·43 .94 .62 · ·33 .76 * Sweets refer mostly to sugar but also include candy and ice cream. Desserts include pies, cakes, puddings. Bread includes bread, rolls, buns, but not toast. For the total group, meat has a significantly higher conflict rating than that of any other food. Its conflict rating, however, varies considerably among the groups, being lowest for the high group and highest for the Czechs and middle group. Although the Czechs are relatively willing to cut meat con
52 The Problem of Changing Food Habits gumption, they show strong conflict in this area. The position of meat in the rank orders of conflict foods increases from third place ire the high group to first place in the low group. Vegetables and milk are second and third highest in the total group. The change in the position of potatoes from seventh in the middle group to second in the low group can be considered an indications of their positions in the diet of these two groups. The great importance of potatoes to the low income group is substantiated by such comments as: "Have to leave potatoes at least once a day," or "If I cooked three meals a day without potatoes (at each meal), there would be quite a fuss." Table ~4 gives the conflict ratings for tl~e total ~,roul~ and tl~e freq~e~cy of the various food categories in answer to each of the three questions re- garding food retrenchment. TABLE 1=ANSWERS OF T~rAL GROUP TO QUESTIONS :E<EGARDING ESSENTIAL FOODS, FOOD RETRENCHMENT. WITH CONFLICT RATINGS Per cent of Per cent of Per cent families Per cent of families Conflict judging it who have families who do not Food rating essential already cut who may cut want to cut Meat ..................... ~.og 75 47 22 24 Vegetables 75 9z ~2 6 40 Milk 62 80 4 x 50 Butter .56 34 I7 6 29 Fruit 54 57 I ~7 26 Eggs 37 5° 8 ~6 Potatoes . 3o 45 9 5 is Bread . 2I 42 7 5 6 Sweets 3~ 6 17 / 4 Desserts ~9 3 ~6 ~ The three foods which produce the greatest conflict, namely, meat, vege- tables, and milk, are also those which are considered most essential. Meat has been by far the most frequently cut food. Although it is considered an essential food, it is one of the most expensive, and cutting it could produce a greater saving than cutting any other food. Only those foods which are considered to be essential are among those which the people say they do not wish to cut. Those which are not considered essential, such as desserts and sweets, are mentioned by very few in answer to this question. D. Roles Played by Uarious Foods The description of the roles played by each food is based on act analysis of the answers to all the questions.* Space permits mentioning here but a few of the foods: Meat fulfills many roles in the American culture. It is almost always considered as a food for dinner, no matter whether the function of the meal * It should be emphasized that these descriptions cannot be considered as complete. Nevertheless these descriptions are an indicator of the kinds of information that can be obtained.
Food Habits and Methods of Change 53 is to be filling, tasty, fussy, healthful, or economical. It is a favorite and essential food for all groups. It is regarded as a special food for the husband by all groups except the Negro group. Meat is the food most frequently mentioned by all the groups as cut due to price increase. Meat is not thought of as a food for lunch when eaten at home. Milk-ranks second to vegetables as an essential food for the groups as a whole. None of the income groups wishes to-cut it even though prices are rising. It is considered a breakfast and luncheon food by about one-third of the people. Milk is not a food that people are cutting. It is not thought of when con- sidering "fuss" occasions and is rarely mentioned as a taste favorite or a filling food. Milk is not a special food for the husband. Potatoes-are accepted by all groups as a food for dinner. They are an essential food and are also thought of as a filling food. The three lower income levels (low, Negro, and Czech) tend to regard potatoes as part of a "fuss" meal more frequently than do the two higher income levels. Potatoes are rarely thought of as a food for breakfast or as a luncheon food. They are never mentioned as a favorite. Salads are thought of most frequently as a food for dinner. They are next most frequently seen as appealing to the taste and as appropriate to a "fuss" dinner. The high and middle income levels consider them as a food for lunch, but the three low income levels seldom mention them as a luncheon food. In general, salads play a more prominent role in the diet of the two highest income levels than of the others. Salads are not considered an essential food, though they are mentioned by about one-fourth of the people as a healthful food. They seldom appear as a special husband's or children's food and are never mentioned as a breakfast food. E. ~ pplication to Problems of Change How strong the forces are which resist changes of food habits in a certain direction can be investigated finally only by actual attempts to change food habits, that is, by an experimental approach. No amount of questionnairing can be a substitute for experiments. However, much of the data gathered from the interviews can be helpful in planning experiments. We will discuss first the results of the specific question concerning sub- stitute foods and then summarize the findings related to the problem of change. i. Substitutability of essential foods. The effect of certain motivational forces toward changes in food habits will depend upon the flexibility of these habits. One factor related to flexibility is the degree to which food which becomes undesirable or unattainable can be replaced by another food. We approached this question by asking the housewives what they would substitute for each of the foods listed as essential. No instructions were given them as to the number of substitutes they should mention, but they were asked
54 Tl~e Problems of Changing Food Habits what they would substitute for each of the essential foods they named. In general the substitutes fall into nutritionally similar categories: oranges for lemons, fats for shortening, cheese and eggs for meat, oleomargarine for butter, another kind of vegetable for the one named, fruits for vegetables, etc. Nutritionally dissimilar substitutes were mentioned only by those in the low income group. Examples of these are shown in Table x 5. TABLE 15 ESSENTIAL FOODS WITH NUTRIT:rONA=Y DISSIMILAR SUBSTITUTES Essential food Eggs ................ Milk ....... Vegetables ...... No. times mentioned No. times mentioned Substitute A specific kind clef cereal 3 Fruit j uices Gravy Table ~5 indicates that the degree of difference between foods which can still be substituted for each other is greater for the low than for the high in- come group. This is in line with the fact that the lower the level of satisfaction of a need the greater is the range of possible consummatory actions for it. a. Basis of change of food habits. Changes in availability of food is one obvious cause of changes of food habits. The area of available food may shrink considerably, as is the case in a situation of shortages. This necessitates a change in type and frequently in amount of consumption. A second cause of changes of eating habits is a change concerning the food channels. An example of shifting to more available channels at the present time is the change to gardening and canning. A third possibility is a psychological change: a food that had been con- sidered "food for others, but not for us'' may become "food for US." Food shortages may facilitate such change. An example is the increased use of glandular meats since the beginning of rationing. Whereas a housewife might heretofore have passed them by, she may now consider them seriously and buy them frequently because of their availability and low "point cost." Similar changes can occur with respect to patterns of meals. In the American culture the "food basket" has three distinct parts assigned to breakfast, lunch, dinner; many foods are considered fit for only~one part. In case of food shortage this might change. Since lunch is the least structured meal (see Table 6) there might be a greater readiness to change the content of the "lunch" than of the other meals. A fourth possibility for change in food habits is to change the potencies of the~fra~es of reference.`This can be accomplished in one of two ways: I) Changing the relative potency of the frames of reference. For example, the current emphasis upon nutritional eating has been planned to increase the relative potency of the "health" frame of reference. ("Eating well to make a strong nation.") 2) Changing the content of the frames of reference, that is the foods related to them. At the present time (May, ~943) the position of fowl has undoubtedly changed from that of a "fuss" food, in the direction of an everyday substitute for other meats which are less available. It is quite
Food Habits and Methods of Change 55 possible that there was some resistance at first to using it as an "ordinary" meat for everyday meals because of its high position in the "fuss" or "com- pany" frame of reference. A fifth possibility for change is a change in belo~ngingness to "eating groups." Increased incidence of school luncheons and eating in factories should be mentioned here. II. EXPERIME~-TS IN CHANGING FOOD HABITS Limitations of time permitted only a few preliminary experiments with changes in food habits. Obviously a great variety of methods can and have been used to bring about change in food habits. They vary all the way from an elaborate treatment of the individual (for instance, in cases of problem children) to mass propaganda by radio, newspaper, billboard, etc. We have chosen for our study techniques which lie halfway between the individual and the mass approach. They are intended for face-to-face groups with an optimum size of ~5 to 50 persons. The approach is not so impersonal as ad- vertisement but lends itself well to relatively quick application to fairly large numbers. Various degrees and types of pressure can be used in the attempt to in- fluence other people. Straight orders with the threat of severe punishment may stand on one end of tl~is continuum free choices on the other. In addi- tion, changes may be brought about by a frank, open approach or by "manipu- lating" people, putting up what actually is a false front and tricking them into certain behavior. "Democratic" methods in regard to changes of groups have been frequently attacked as being rather wasteful. We have been particularly interested in trying out a method which we have named "group decision." A group de- cision has one aspect in common with group discussion; namely, that a free interchange of ideas takes place and that in many respects the initiative lies with the group. No attempt is made to force a decision on the group or to use high-pressure salesmanship. On the other hand, unlike a mere group discus- sion, group decision leads to a setting up of definite goals for action. These goals may be set up by the group as a whole for the group as a wholes or by each individual in the group setting for himself.3 Unfortunately there is no space here to give the necessary details of the method. It may be stated that it is by no means simple to lead a group discussion within forty-five minutes to a conclusion without "manipulation." The procedure obviously requires very thorough considerations in regard to the creation of a proper atmosphere, a definite leadership, and the use of experts in the right way at the appropriate time. It is planned to give a full account of these experiments elsewhere. At. Group Decision and Request as TIIea~s of Changing Food Habits * 6 This first experiment attempted to determine the relative effectiveness of two methods for changing food habits of a group: ~) Group decision the *This experiment was conducted by Ben Willerman.
56 The Problem of Changing Food Habits group decides for itself whether and to what degree it wishes to change its habits. z) Request the group is requested to make a change to a certain extent. I. Sub jects and procedures. The experiment was concerned with the increase in consumption of whole wheat bread as compared with white bread. It was conducted during April and May, ~942. Eight cooperative dormitories for men students at the State University of Iowa were used. Their membership ranged from no to 44 men; the average age of the students was approximately Hi years. The eight co-ops were grouped into four pairs, each pair being matched on the basis of similar percentage consumption of whole wheat bread at the start. One of each pair was to make a group decision and the other was presented with a request. Fo'r a week preceding the experiment and during the experimental period of one week, only whole wheat and white bread were served. During these two periods daily measures of consumption of both kinds of bread were made by student waiters. They were instructed to keep the dining-room tables supplied with both kinds of bread throughout the meals so that the members could choose freely. The student proctor of each of the dormitories understood beforehand that he would receive a letter by an authority to be read at the end of the meal when all members were present. After reading the letter which instigates "group decision," he would ask for discussion of the proposal. If the group agreed to cooperate in the experiment, he would ask them to decide how much they would increase their whole wheat consumption for the following week. The "request proctor" was merely to read the request letter and ask for comments. All groups were informed of their previous average percentage consumption of whole wheat. The amount of change desired of a particular request group was set at the since level as the amount which had been voluntarily chosen by that decision croup with which it was Haired. A description of the meetings was ~ , · ~ ~ ~ A ~ . . · · . ~ · ~ ' ' obtained by an observer. Atter the experimental period the members were asked to fill out a questionnaire about their bread preferences, and about va- rious aspects of the experiments. 2. Results. a. The level of the group goals. All group decision co-ops voted for an increase of whole wheat consumption. From their previous levels of ap- proximately solo, one group decided to go to the 669fo level, one to Hobo and two to loom. Of the two groups which voted for loom, group C was very enthusiastic over the proposal and quickly and unanimously voted for that goal. Group D was fairly evenly divided between setting a loom goal and setting a lower goal. The loom level was decided upon by a small majority only after a
Food Habits and Methods of Change 57 rather bitter fight. The results of the questionnaire indicate in a striking wall the effect of this disagreement. The two request groups who were asked to raise their whole wheat con- sumption to loom did not resist this imposition of an extreme goal. b. The effect of group decision and request. I. Change in consumption. The effect of the group decision as compared with the request can be measured by the difference in the amount of actual change in consumption and by the attitudes of the group members as revealed by the questionnaire. Various circumstances interfered with the getting of satisfactory records of consumption in this experiment; for example, the group that had volun- tarily chosen to increase whole wheat consumption to hobo advised the waiter to serve but a few slices of white bread. This was an expression of their eagerness to reach the group goal but prevented a scientific comparison. Other groups did not have sufficient bread of one or the other type delivered by the bakers. From the data available we observe a superiority of the group decision co- ops. The group decision co-ops that set their goals at 6670 arid god reached their goals, whereas the parallel request co-ops did not. a. The in~rnediate reaction to the proposal. The questionnaire which was filled out after the experiment by those students of the co-ops who had not left for vacation contained two questions related to their immediate reactions to the proposal. Question a): "When the proctor read the letter announcing the experi- ment, what was your reaction to it?" Of the request groups (Table ~6), about half were favorable (487o); the rest mainly indifferent. Of the group decision groups (excluding Group D), the great majority of the members (7870) were favorably inclined. Only ~ were indifferent and only 37o unfavorable. Group D, however, which decided by a very small majority on loom whole wheat consumption, had more than half of the members reacting unfavorably. TABLE 1WREACTION TO PROPOSAL TO PARTICIPATE IN TEE Favorable Request Co-ops (E, F. G. H) N ~9 % 48% Group Decision Co-ops Indif- ferent I6 4 EXPERIMENT Unfavor- Miscel- able laneous 2 3 5~0 8% (A, B. C) N 28 4 I 3 70 78% I I ~0 3% 8~0 (D) N 5 4 I ~° % 2470 19% 57% 070 N = number of persons participating in the experiment. On the whole, we can say that a group decision to change food habits is more favorably. accepted than a request from someone outside the group. However, a decision with a very small majority led to a reaction which was decisively less favorable than the reaction to the request.
58 The Problems of Changing Food Habits Question b): "Did you feel that the goal set for the group was just right, too high, or too low?" elicited similar reactions from both the members of the request and group decision co-ops. The great majority (8070 of request groups and 8670 of group decision groups) felt that the level set was right. This seemed to be independent of whether the goal was set relatively close to or far above the previous consumption. Group D which had made its decision by a small majority shows also in this question its discontent: most of the members (7870) considered the goal too high, but the remainder (~) said it was "just right." 3. Eagerness to reach the goal level. One question was asked referring to the person's own eagerness to succeed, another referring to his judgment about the eagerness of the group as a whole. Ouestion a): "How eager were vou for the croup to reach its coal?" J ~ O~ ~-~ --- ---- ~ r -~ -em-- --~ ~~ - - ~. . . ~ . ~ . Ha. . ~ ~ ~ l his was answered on a scale ranging from a plus-4 to a m~nus-4. l he middle point, c, meant: didn't care one way or the other. The members of three of the group decision co-ops rated themselves more eager to succeed than did the members of the parallel request co-ops TABLE 17 EAGERNESS TO SUCCEED Rating of own eagerness Rating of group's eagerness ~ it_ Goallevel Gr. dec. RequestGr. dec. Request 6670 2.5 I.9 2.4 I.8 90% 2.8 2.0 2.6 I.6 loo% 3.6 ~.6 2.8 T.8 Average z.9 z. ~.6 ~.7 IOO~o I.0 2.0 -0.4 2.0 (Co-op D) (Co-op D) (Table ~7~. The only exception was again Group D. The members of this group rated their own eagerness lower than any other group. Question b): "How eager do you think the group as a whole was to reach their goal?" With the exception of Group D the members of the group de- cision co-ops judged their own group to be more eager to succeed than did the members of the request groups. Group D again showed by far the least eagerness to succeed. In fact, they expressed on the average the opinion that their group had a slight wish not to succeed (Table ~7~. On the whole, then, we can say, if we disregard Group D, that all co-ops were satisf ed with the level set for the change of their food habits indepen- dently of whether this goal was introduced by group decision or request and independently of whether it was high or low. The members of the group d~ecisi~ co-ops showed more favorable attitudes to the proposition as a whole and were more eager to succeed. In one co-op where group decision was based on but a slight majority, the attitude was decidedly unfavorable. There was little or no eagerness to succeed and the goal level was considered too high. As a point of general interest it might be mentioned that the individuals rated their own eagerness to succeed higher than that of the eagerness of the other members of the group. This holds true for both the group decision co- ops and the request chops with the exception of one request co-op where the rating was equal (Table ~7~.
Food Habits arid Methods of Change 59 4. Food preferences arid eagerness to succeed. For members of the request co-ops there is a definite relation between their eagerness to succeed and their personal preference for whole wheat as against white bread (Table ~8~. The more they preferred whole wheat, the more eager they were for the group to succeed. The members of the group decision co-ops did not show this rela- tionship. Independ;ent of their personal likes and dislikes, they were equally eager to see the group succeed. TABLE 18 RELATION BETWEEN ORIGINAL PREFERENCE FOR WHOLE WHEAT BREAD AND EAGERNESS TO REACH GOAL Very much Request Co-ops ~.4 ( i4) Group Decision Co-ops ( A, B. C) 3.2 (5) Group Decision Co-op D (o) Prefers whole wheat MUCh 2.3 ( I4) 3.2(II) 37 (3) The numbers refer to average ratings of preference. The numbers in parentheses refer to the number of individuals. No prefer ence I.g(7) 3 3 (7) I.I (7) Prefers white bread Much Very much I.3 (6) 0 (I ) 3.3(IO) T.5 (4) I.7(3) -2.0 (6) This indicates tliat in case of an external request the wish to succeed or not to succeed depends largely upon whether or riot the personal likes happen to be in line with the request. ~ group decision, on the other hand, seems to establish a sufficiently strong group goal to be accepted by the mem- bers in a way which overrules to a considerable degree personal taste. The motivation induced by group decision is higher than that of the re- quest groups in spite of the fact that, in our sample, the preference for whole wheat is greater in the latter (average of 3.7 as compared with 2.9 on a 4-point scale). That the psychological acceptance of the group rather than the form of group decision is the decisive factor is illustrated by Group D which had a small majority decision. Those members of that group who preferred white bread very much had a strongly negative attitude toward reaching the goal. c. The eliect of increased consumption oil preference. The increased con- sumption of whole wheat bread during one week seemed to have led to a slight increase in preference for this food. Those members whose ratings could not increase (already had a rating of "prefer whole wheat very much") or decrease (had rating of "prefer white bread very much") were omitted from the tabulation. Of the others in the group decision co-ops, 2370 in- creased their preference for whole wheat, while oh decreased. In the re- quest co-ops, 2670 increased their preference, and 27o decreased. Group D, with the small majority decision, shows the opposite trend, with 5~ increase and 20~ decrease. d. Erect of goal levels on attitudes toward future consumption. There seemed to be a difference in the attitude toward future consumption of whole wheat bread between those groups whose experimental goal was be- low loom and those on the loom level. Eighty-three per cent of the former groups wished to remain at or increase their level whereas only ~8~ of the latter groups wanted to maintain their loom level. A bread diet of only whole wheat seems to lead to psychological satiation and a desire for variety.
60 The Problem of Charging Food Habits 3. Summary. A method of group decision is compared with a method of request in its effect on changes in food habits and related attitudes. The group decision method seems to create a snore favorable attitude; the individuals are more eager to succeed; and their wish to cooperate is more independent of their personal likes and dislikes. If the group decision is based upon too small a majority there is danger of a "kick-back" which seems to make the outcome even less favorable than that of the request method. B. The Relative Effectiveness of a Lecture Method and a Method of Group Decision for Changing Food Habits * 5 i. The problem. The interest in the group decision method stemmed from the realization that even strong motivation may not suffice to change the action of a person in regard to his food habits. Motivation leads to action only if it is brought down from the level of wishes and sentiments to the level of a "decision." This experiment was an attempt to evaluate the effectiveness of a group decision procedure, in which a nutrition expert cooperates with a group- discussion leader. The expert provides the technical knowledge in a social setting which induces decision. The change in food habits instigated by the nutritionist in a group decision setting was compared with the effect of the same nutritionist in a lecture setting. a. Experimental procedure. a. Length of Sleeting. Since practical usefulness of any procedure for changing food habits on a large scale demands that the method riot be too elaborate or time consuming, both the lecture and the group decision pro- cedures were limited to about thirty minutes. b. Food: changes attempted. It was thought that attempting a rather diffi- cult food habit change would be a more severe test of the method, thus per- mitting safer conclusions. A preliminary experiment which attempted to introduce a new vegetable (escarole) showed the feasibility of the group- discussion method. In the present experiment kidneys, brains, and hearts were chosen because of the known resistance to these foods. There are defi- nite indications that this resistance is a rather deep one, frequently combining elements of physical aversion, social status, and superstition. c. The groups used. To permit comparison of the method of change, two groups of women were selected from each economic level, one for the lecture method and one for the group decision method. One pair of groups came from the best residential areas, one from substantial middle class homes, and one from a lower economic stratum but still above subsistence level. * Indebtedness is expressed to the Cedar Rapids Red Cross, the Cedar Rapids Nutri- tion Service and the Home Economics Department of the State University of Iowa for their cooperation and assistance in the conduct of the study.
Food Habits and Methods of Change 6I All six groups, ranging in size from IS tO I7 members, had already been organized for Red Cross Home Nursing courses and met regularly at private homes or schools for the Red Cross course. d. Experts and group leaders used. Mrs. Mulqueen, who has been the city nutritionist in Cedar Rapids, Iowa, for several years, and Mrs. Hubbard, who is recognized as a leading American volunteer leader and organizer in the field of nutrition, both functioned in some of the lectures and group de- cisions. They were selected so that a comparison could be made between the group decision procedure and superior nutrition lectures. The leader of group decision was Mr. Bavelas, who has had considerable experience in group work. 3. Procedures used irz the lecture and group decision. Stenographic records were made of some of the lectures and of the group decisions. Briefly, the procedure in each case was as follows: a. Lecture. The lecturer linked the problem of nutrition with the mar effort, emphasized the vitamin and mineral value of the three meats giving detailed explanations with the aid of charts. Both the health and economic aspects were stressed. The preparation of these meats was discussed in detail as well as techniques for avoiding those characteristics of these meats to which aversions were oriented (odor, texture, appearance, etc.~. Mimeographed recipes were distributed. The lecturer was able to arouse the interest of the groups by giving hints of her own method for preparing these "delicious dishes," and her success with her own family. b. Group decision. The group decision method is based on a cooperation between the group leader and the expert. It should be emphasized that no attempt is made to "high-pressure" the group into any kind of "promise" to serve these foods. The group-discussion leader starts with a very short introduction linking the problem of nutrition with the war effort and with general health. He points to the difficulties the government has met in trying to change food habits, and asks the opinion of the women, as a representative group of housewives, as to how successful a direct appeal to groups of housewives like themselves would be. From here on, the group is led step-by-step to seeing the problem more concretely, and, at the same time, to taking over the re sponsibility "to do something about it." This point is reached through a free group discussion which also brings out the specific reasons why the meats have been rejected by the housewife herself or by her family. These specific objections are the cue for the introduction of the expert, who discusses the various methods of getting around the difficulties. On the whole, the same information is imparted by the nutritionist in this setting as in the lecture, but in a condensed form (7-~o minutes) and the same mimeographed recipes are distributed. A "census" taken in the earlier part of the meeting has shown how many of the women have served any of these foods in the past. Now the group is ready to vote on the question of trying one of these meats in the following week.
62 The Problem of Charging Food Habits 4. Results. a. Collection of the data. The effect of the lectures and group decisions were tested after seven days. An interview at home with each participant was used to determine the extent to which the foods were served and what the reaction of the family was. b. Frequency of servings of heart, kidney, and brains' after lecture and after group decision. Before the experiments, the frequency with which these meats were served by the participants in the lectures and the participants in the group decision were about equal. In both cases, kidney and brains were very infrequently served. The results of the experiments are shown in Table ~9. T A B L E 19 C MANGE OF FOOD H ABITS Economic level .......... Number of participants.... .............. Low ............. . ~7 Per cent of individuals serving one or more of the three meats 35 Per cent of individuals serving a meat they had never or hardly ester served Group decision Lecture it, Middle High Total Low Middle High Total ~6 13 44 IS ~5 :3 4I 6g 54 52. IS IS 0 To - before ~O 53 54 44 O 8 o 3 Per cent of individuals serving a meat they had never served before 13 36 50 32 O S O 3 ~ , . . . . Her cent of partlapants serving one or more new meats who had nearer served any of the three meats before the experiment ........................ . . . Out of total of r4 participants. Out of total of ~ I participants. .. fig* .. Of Within the seven-day period after the lecture meeting, four women, out of a total of 4~ participants of the lecture group, served at least one of the three meats. Out of a total of 44 participants in the group-decision groups women served at least one of the glandular meats. In other words, the lectures led to action by logo of the participants; the group decision led to action in s27o. Furthermore, after group decision, 2370 of the women served a food they had never served before. whereas the corresponding nercenta~e for the lecture group is 370. rid ~, ~ --<, i: - - in,= 1ne same trends are shown it we consider each economic level separately: the percentage of women who served a food that they had never served before varies between To and hobo for the participants in group decision. For the lecture group, it is zero ~ for two groups and 8.3~ for the other group. Fourteen participants in group decision had never used any of the three meats before; this indicates that the food was highly negative to them. Twenty-nine per cent of these individuals tried one of these meats in the following week; whereas none of the eleven members of the lecture groups who had never tried any of the meats used them during the seven-day period.
Food Habits and Methods of Change 5 Summary. 63 An experiment comparing lecture and group decision methods for change of food habits was carried out with about one hundred and twenty women from varying economic levels in a medium size town in Iowa. The check-up after seven days showed that when the nutritionist functions in a group decision setting, the changes are definitely greater than when this same nutritionist functions in a lecture setting. The success with glandular meats indicates the usefulness which this method may have even in cases of strong resistance to changes in food habits. This procedure is in line with recent attempts at group education in various fields: the individual is approached in a group setting, and demo- cratic discussion rather than lecture is stressed. No attempt is made to manipulate the group by high-pressure sales talks. It should be emphasized that the procedure described here is not merely a group "discussion," but a discussion leading to a decision. In this particular version, this decision has not the characteristics of a "resolution" of the group, although there may be cases where such a decision for group action is advisable. Rather it is a decision made by the individual concerning her own action the housewife decides what she will do at home. The group setting gives the incentive for the decision, and facilitates and reinforces it. C. Relation of Group Decision Method to Findings of Interzrzew Study The procedure reported in Section B contains the following steps (de- scribed in greater detail in the mimeographed report): ~) setting the stage; 2) eliminating the possible implication that the problem is typical of a low ~ . ~ ' · ~ 1 ~ _ 't ' 1 · ~ _ _ _ ~ ~ ~ A: _ ~ status group; 3) request for group opinion on the pOSSlDlilty OI changing food habits; 4) discussing the particular foods to be changed and taking the census; 5) the nutrition expert answering questions in regard to difficulties of preparation, etc.; 6) asking for a group decision; 7) discussion of method of presenting the food to the family; 8) discussion of fleterminin~ the suc- cess of carrying out the decision. i, The content of the discussion seems to have been guided by the same con- siderations which come up in the interview data. These data suggested that food habits might be changed by changing, for instance, the food channels, the place of a particular food inside or outside the area of consideration as "food for us," the frame of values governing food conduct, the high or low position of a given food on the various value scales. I. When the discussion leader dealt with the glandular meats in terms of health food, they were automatically placed in the area of consideration as "possible foods for us." For those women who were already somewhat familiar with the health value of this food, its position on the "health scale" was raised. 2. Greater potency was given to the health area by linking health with patriotism. This was accomplished by the explanation of the Government's concern over the number of young men rejected by the army on grounds which might be related to nutrition.
~4 The Problem of Changing Food Habits 3. The experimental foods, particularly kidney, were placed on a higher status level and perhaps even on a "fuss" level by describing the widespread use of kidney in England. 4. The conflict associated with "low-money food" was removed by combin- ing emphasis on the inexpensiveness of the glandular meats with emphasis on "health" and "status." 5. The obstacles in the path to the goal were removed, either in the dis- cussion or by the comments of the nutritionist. Some of the principal obstacles were: a. Buying. This channel was clarified; the women were told where they could get the glandular meats, that they would have to tell the butcher ahead of time when they wanted them, etc. b. Preparation. One outstanding objection raised by a number of the women was the bad odor of kidneys during cooking. A way of overcoming this and other difficulties were described. Recipes for each of the meats were distributed. Presentation to the fancily. The women discussed how they might best present the new foods to their families. (For instance, one woman reported that she overcame her husband's objection by telling him what it was only after he had expressed his liking.) 6. The group decision had a "freezing" effect for future action. This is an aspect missing in the lecture method. After a lecture many paths are still open whereas after a decision the person has committed himself to follow one path. After the previously existing food ideology had been loosened up in the discussion period and after forces had been created which lead to changes, the decision is the first step toward stabilizing the food ideology in the new pattern. III. Sum MARY This study deals with the problem of why people eat what they eat arid with certain experiments in change of food habits. Food behavior is determined by the dynamics of the food situation which includes the channels through which food comes to the table, the gatekeeper governing the channels at the various points, and the food ideology of the gatekeeper. A system of values is the basis of some of the forces which determine decisions about food and bring about conflicts of varying intensities. In the experiments on changes of food habits, a group method, group de- cision, was compared with a lecture method and request These experiments suggest that it may be possible to bring about in a relatively short time defi- nite changes in food habits, even in food items which would be expected to show great resistance to change. BIBLIOGRAPHY I. Bavelas, Alex. A method for investigating individual and group ideology. Sociometry, 5 :37~-377, Ugly. 2. Engel-Frisch, Gladys. A study of the effects of odd-shifts upon the food habits of war workers. Washington, D. C., Committee on Food Habits, National Research Council, August ~942. 3 p. Mimeographed.
Food Habits and Methods of Change 65 3. Kalhorn, Joan. Ideological differences among rural children. In Lewin, K., ed. Studies in topological and vector psychology III. Iowa Studies in Child Welfare. (In press.) 4. Lewin, Kurt. A group test for determining the anchorage points of food habits. Washington, D. C., Committee on Food Habits, National Research Council, June 19~. 21 p. Mimeographed. 5. Lewin, Kurt. The relative effectiveness of a lecture method and a method of group decision for changing food habits. Washington, D. C., Committee on Food Habits National Research Council, June ~942. 9 p. Mimeographed. 6. Willerman, Ben. Group decision and request as means of changing food habits. Washington, D. C., Committee on Food Habits, National Research Council, April ~943. ~ p. Mimeographed. 5
