K'ELATEQNEHIP’S l-ETWEW RGLE AND SELF PEKCEP‘YIGNS, FAMILY CHARACT‘EREfiTECS‘ SHGF‘P’ENG ATTITLIIDi-ES AND FQQD f3EiRCHASE BEHAWQR “tests £01? {'56 Degree of M. S. MHCHIGAN STATE UNIVERSETY David Ho-Ifiiagswurth Spaeth 1960 y ,3 E 35% a ». ‘ 76:11.“? . $ 1“,, a x I ENTARY g 't' '1 '3 .7: {‘4' W é m 33ml i K OF ’73?" ' im‘k‘fin w: ~ g u wager/g. is " .3 ea. .. A- \ A "'\ P. '1‘ 1 3 3' '9 I 1“; \H W. . u—..Ii. U i :‘_ d H ’B\ 3 LIBRARY Michigan State University J RELATIONSHIPS BETWEEN ROLE AND SELF PERCEPTIONS, FAMILY CHARACTERISTICS, SHOPPING ATTITUDES AND FOOD PURCHASE BEHAVIOR BY DAVID HOLLINGSWORTH SPAETH AN ABSTRACT Submitted to the College of Agriculture of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Economics 1960 .r’? i I, [O I 7 If} 1' ’I .4 Approved K [,5 .7 ibg$l‘£i/" ' 1/", g} ,1: ’1 ‘4.” .’ ,W'— .1 I f,» // " x (L 2 David Hollingsworth Spaeth ABSTRACT The objectives of the study were: 1. To find whether or not statistically significant relationships exist among observed shopping behavior, verbalized attitudes toward shopping, role and self perceptions of the homemaker, socio-economic characteristics of the family, and the expenditures for or quantities of specific foods purchased. 2. To explore the notion that the decision-making process is a user of a finite quantity of energy such that the decision-maker finds , it necessary to ration this energy among decisions. The existence of a biological relationship to this energy was also hypothesized. 3. To demonstrate the usefulness and shortcomings ofabivalued cross-sectional, analytical procedure in dealing with intercorrelated survey data and variables not quantifiable in a cardinal sense. The investigation proceeded as follows: 1. A sample of 211 Lansing families was investigated. These families were members of the Michigan State University Consumer Panel, reported their food expenditures and other data each week for 52 weeks in 1958, and responded to a special questionnaire dealing with role and self concepts and buying attitudes in January, 1959. 2. Each family was described as being in a high or low value class for each of variables considered to have relevance to the decision-making process in shopping for food. 3. These variables were paired in all possible combinations by means of 2. x 2 tables of frequency and tested for significance. 4. Differences in the use of specific replies to the role- self questionnaire were recorded at the . 05 level of significance. 5. Socio-economic subgroups were defined within the 211 family sample and a specific hypothesis relating role—self conflict scores to,(a) food purchasing behavior,(b) attitudes toward shopping for food and (c) per capita expenditure for food was tested. A second hypothesis was tested in the subgroup analysis relating per capita milligrams of thiamin in the food purchased to, (a) role-self conflict score, (b) attitudes toward shopping for food, and (c) food purchasing behavior. In the analysis of all possible pairs of variables signi- ficant relationships were found among observed shopping behavior, attitudes toward shopping, role and self—perceptions of the home- maker, socio—economic characteristics of the family, and expen- ditures for or quantities of specific foods purchased. Expenditure levels for a number of food groups were found to be associated with per capita income, percent of income spent for food, and variety of food purchased in terms of per capita number of different items purchased and percent of total different items purchased only once during 1958. Role-self conflict score was related to attitudes toward shopping for food, food purchasing behavior, per capita income, family composition and per capita milligrams of thiamin in food purchased, but not to the food expenditure variables. Per capita milligrams of thiamin in the food purchased was related to role-self conflict score and attitudes toward shopping, but not to the shopping behavior variables. A complex of consistent relationships was observed in ”which a homemaker with a low conflict score also tended to show low orientation toward price calculations, included relatively few items in 50 percent of the food bill and purchased a low level of thiamin. This same homemaker was also under income restriction and pressed for time in the home. The association of low conflict score with these variables supported an inference that the low role- self conflict score is a result of a low energy availability and pre- occupation with household duties that compete with food purchasing activities for the homemaker's time and attention. These observations lend support to a hypothesis that a rationing of energy among decisions may take place and that it is associated with the physiology of nutrition. RELATIONSHIPS BETWEEN ROLE AND SELF- PERCEPTIONS, FAMILY CHARACTERISTICS, SHOPPING ATTITUDES AND FOOD PURCHASE BEHAVIOR By DAVID HOLLINGSWORTH SPAETH A THESIS Submitted to the College of Agriculture of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Economics 1960 ii ACKNOWLEDGMENTS I wish to express my appreciation to the Department of Agricultural Economics for the means and the academic freedom to pursue this course of investigation. I am particularly grateful to Dr. James D. Shaffer for the rather liberal use of his time spent in the informal conversation necessary to the development of this kind of work. iii TABLE OF CONTENTS Chapter Page I. OBJECTIVES AND APPROACH .......................... 1 Objectives of the Study 1 Method 3 The Decision-Energy Concept 5 Procedure 9 Definitions and Derivations of the Variables 12 Tests for Significance 22 Assumptions in the Study 26 II. RESULTS OF THE ANALYSIS OF THE FULL SAMPLE ..... 2.8 Per Capita Income 33 Per Capita Expenditure for Food 34 Age of Homemaker 37 Education of Homemaker 40 Number of Meals Served at Home 40 Presence or Absence of Children, Age 0-9 41 Presence or Absence of Children, Age 10-19 45 Percent of Total Number of Different Food Items Purchased Only Once . 46 Number of Different Items Purchased 49 Percent of Income Spent for Food 51 Number of Items Included in 50 Percent of the Food Bill 54 Per Adult Expenditure for Coffee 55 Per Capita Expenditure for Sugar and Confections 56 Per Capita Dozens of Eggs Purchased 58 Per Capita Expenditure for Vitamin and Mineral Supplements 59 Per Capita Expenditures for T. V. Dinners 60 Per Capita Expenditures for Dairy Products 61 Per Capita Expenditures for Fats and Oils 63 Per Capita Expenditures for Fruits 64 Per Capita Expenditures for Vegetables 66 Per Capita Expenditures for Meats 67 Per Capita Expenditures for Bakery Goods 69 Per Capita Milligrams of Thiamin in Food Purchased 7O Conflict Score 72 Social Score 74 Price Oriented Shopping Attitudes 76 Convenience Oriented Shopping Attitudes 77 iv Chapter Page III. SUBGROUP ANALYSIS ................................. 80 Relationships within Subgroups 82 Results of Analysis of Subgroups with Respect to 83 Socio-Economic Variables Summary of Findings 87 IV. SUMMARY AND CONCLUSIONS ......................... 90 APPENDIX ................................................... 94 CHAPTER I OBJEC TIV ES AND APPROACH The primary investigation in this thesis is directed toward the question: Does a relationship exist among (a) observed purchasing behavior, (b) verbalized shopping attitudes, (c) role and self perceptions of the homemaker, (d) socio-economic characteristics of the family, and (e) the expenditures for or the quantities of foods or food classes purchased? A second question, which carries with it implications concerning the nature of the decision-making process, asks whether or not decision-making is an energy-consuming process such that the decision-maker finds it necessary to ration his expenditure of this energy among decisions. This study can do no more than suggest answers to these questions within the limitations of the data presented here, but evidence to support affirmative answers to both questions will be presented. A third point of emphasis in the course of discussion is the method used to detect the presence or absence of significant relationships in the data. The content of the primary question above is similar to that of questions faced when coping with problems determining the content and direction of advertising and the marketing of a commodity. An elementary procedure and kind of relationship is the identification of unique sets of characteristics which are characteristic of families at particular expenditure levels for food, and who show various kinds of purchasing behavior. The initial objective, therefore, is to sketch a profile of the families possessing these expenditure and behavior characteristics. Extending the idea of unique relationship further, there is reason for suggesting that there is a biological basis for expenditure levels of food being related to role and self perceptions assuming that expenditure and quantity are correlated and to the extent that role and self perceptions are unique properties of individuals. Such relation- ships are suggested by experimental evidence that the nature of the food eaten by an individual can affect that individual's behavior and attitudes. Two examples of such studies are, first, a U. S.D. A. investigation1 which related level of breakfast protein intake to sub- sequent blood-sugar levels and associated states of fatigue, nervousness, irritability, or, in contrast, feelings of well-being. The second was an experiment in which a group of hospital employees volunteered to live on a diet in which the thiamin content had been reduced to . 22 mg. per day per person, although adequate in other respects. These hospital employees showed marked changes in behavior and personality. 1Orente-Keiles, E. and Hallman, L. F. The Breakfast Meal in Relation to Blood Sugar Values, U. S. D. A. , Circular No. 827, 1949. ZWilliams, R. 1)., Mason, H. L. Smith, B. F., Wilder, R. M. "Induced Thiamin Deficiency in Man," Archives of Internal Medicine, Vol. LXIX, 1942, p. 721. It is not the purpose of this study to deal with nutritional deficiency. Rather, the attempt is to establish whether or not there is statistically significant evidence that the nutritional components of the food purchased can be associated with shopping attitudes, observed shopping behavior, and role and self-perceptions as defined herein. The question of causal relationship must be left to more precisely controlled study than is possible with the data at hand, but a relationship is suggested in the conclusions which is consistent with the findings. The matter of whether this suggested causal scheme is more valid than another, equally consistent, cannot be handled here. The kind of relationships sought to explore this area involved the calculation of the thiamin in the food purchased, per capita, per day, and the comparison of this to a role-self conflict score and various measures of shopping behavior. Expenditures for principal sources of protein were compared with psychological variables in the attempt to find significant relationships. METHOD It is obvious that the relationships that are sought are subtle and that they involve complexes of a large number of variable characteristics. The problems of intercorrelation and nonquantifiable variables had to be faced. To deal with such problems a cross- section type of analysis was used in the following sense. The sample under analysis consisted of families participating in the Michigan State University Consumer Panel. Each family was described with respect to each variable used in this study, e. g. per capita income, per capita expenditures for foods, psychological scores, etc. Each variable was permitted two values. With respect to the variables which were characterized by ordered number scales, each was designated as being high or low (e. g. high per capita income means greater than $2199. 99, low per capita income means less than $2200. 00, after federal income taxes in 1958). In the case of the psychological variables, number systems were devised and high and low classes defined as with the cardinally measurable variables. In the case of describing family composition with respect to children, the classes were simply, present or absent. No claim is made for this bivalued aggregation device other than as a probe for uncovering possible relationships. Rather broad outlines are suggested in the context of discovery and although validation procedures are used, each relationship uncovered should be tested more rigorously and in other populations. The bi- valued aggregation has the very great advantage of simplicity, a considerable feature when handling large numbers of variables simultaneously. A second device is used to provide some insight as to the nature of the value classes for each variable. The frequency with which the value classes for each variable used specific phrases and adjectives in reply to the role-self questionnaire was noted and 3These adjectives and phrases are the replies to questions in parts II and III, 3rd Special Panel Questionnaire, see Appendix p. expressed as a percentage of the number of respondents in the value class. The percentages of frequency for the two value classes for a given variable were compared and significant differences were reported. THE DECISION ENERGY CONCEPT The second question advanced for study deals with the concept of a decision energy. This is the intuitive notion that people generally find it very difficult to deal with two important decisions of divergent content simultaneously. The idea of a finite quantity of a mental energy is not new to psychology. It is implied in a number of theories as an explanation of the motivating force or drive which produces physical activity. In psychoanalytic theory a kind of stable output of energy has been implied such that if its expression in one form is suppressed, it is released in another form of behavior. A current explanation of neurotic behavior employs the idea that neurotic conflict consumes energy required for normal, flexible reaction to en- vironment. The disproportionate use of energy in implementing various defense mechanisms leads to fatigue and/or, the rigid inappropriate behavior characteristic of neuroses. 4 The presence of decision energy in making economic decisions is implied in current views held in theory of consumption behavior. Both the equimarginal principle and the indifference approach to consumer behavior assume that the individual is a precise 4 White, Robert W. The Abnormal Personality. New York: Ronald Press Co. , 1956, p. 257. and sensitive calculator possessed of perfect knowledge and foresight. Stigler, however, comments: The inertia that prevents instantaneous readjustment of the consumer‘s behavior is not necessarily irrational. The consumer may deem the effort to readjust his purchases excessive in the light of probable reward--decision-making is not a free good--so he develops stable patterns of be- havior (habits) in dealing with the relatively stable environ- ment in which he lives. 5 Shaffer, in describing habit, states: In a modern economy, the number of alternative commodities which might be purchased is so great that it is impossible for one person to consider more than a small fraction of them. The consumer must develop techniques for making consumption decisions which will economize time and effort. Such statements imply that a person tends to make one decision at a time and that the process exacts a time and energy cost. The pervasive use of habit is more striking if one tries to imagine the overt behavior of an individual devoid of reliance on habit, who pre- ceded each action with a specific deliberated decision. It is seen that the energy concept leads to consideration in detail of habitual behavior. A first definition of habit may be repeated performance of an act, such as, the repeated purchase of a specific brand of merchandise. This is not adequate since the price and other knowledge concerning the article in question may be such that the 5Stigler, George J. The Theory of Price. New York: The Macmillan Co., 1957, p. 46. 6 Shaffer, J. D. Notes for Agr. 240, Chapter 8. Vincent (Ed. ), Department of Agricultural Economics, Michigan State University, 1959, Mimeo. repetition would be indicated under a behavioral model as sensitive as that assumed under the indifference or equimarginal approaches. What is really meant by habit is then a certain insensitivity to change resulting from an unwillingness or inability to reassess reality. Schreier asserts that observed repetition is not a sufficient basis for prediction and comments on habit as follows: Repetition of an action . . . , is not the criterion of a habit . . . . The statement that "he has a specific habit" means that he has acted by making certain assumptions which are reasons for his action; but only if we include in our prediction that the individual will continue his assumptions and not reexamine them can we predict that the individual will continue to act in this way, repeat the habitual action. 7 . . . . ,8 ,, Similar notions are advanced as Norris "petty goods category of purchases, Katona"s9 "genuine decision” and Schreier's "founding decision. " The investigation of habit in this study is pursued by using certain shopping attitude variables as measures of willingness to re— assess reality and comparison of this to variables which may be logically construed as measures or determinants of energy. The variables in this study that are directed toward manifestation of habitual actions 7Schreier, Fred T. Human Motivation. Glencoe, Illinois: The Free Press, 1957, pp. 84-85. 8Norris, Ruby Turner. The Theory of Consumer's Demand. New Haven: Yale University Press, Rev. Ed., 1952, p. 100. 9Katona, George. Psychological Analysis of Economic Behavior. New York: McGraw-Hill Book Co. , lst Ed. , 1951, pp. 49-50. 10 Schreier, op, cit.., p. 83. are concerned with identifying patterns of repeated purchases. These variables are (a) number of different items11 purchased during 1958, (b) percent of the number of different items purchased during the year 1958 which were purchased only once, and (c) the number of different items which were included in 50 percent of the food bill. These variables were coupled with verbalizations concerning attitudes in shopping which were oriented toward price and convenience. These specifically oriented shopping attitudes were derived from questions implying a willingness to spend time in food preparation and engage in associated calculations and decision-making activities. 12 The decision energy concept is further explored by searching for relationships between these aforementioned variables and the conflict between role and self-perceptions. The nature and import of this conflict will be discussed in defining the conflict score variable. The idea is, that if it is true that an energy rationing is operative, and that conflict between self and role perceptions are associated with the energy available for decision-making, then it should be possible to observe a set of relationships that will be consistent within some scheme of causality. 11 . . . . Different item means a separate food item as indicated by a different Consumer Panel Diary code number. See Appendix 12 . . . The meaning of these shopping attitudes may best be seen by examining the questions used. These are listed in Appendix p. 95. PROCEDURE The reader will quickly recognize that certain of the variables are highly interrelated, such as, per capita income with percent of income spent for food. It may be seen too that examining pairs of variables by means of 2 x 2 frequency tables may be misleading to the extent that a lack of relationship may arise from antagonistic effects from subgroups within the sample. An example of this would be contrasting attitudes toward price as expressed within a low per capita income class consisting of both young couples with no children, and older couples with no children but the latter possessed of habits of thrift ingrained by the rigours of raising a family and lower expectations of future income. In this case the group must be further subdivided by introducing the age of the homemaker as another relevant variable, in order to prevent cancellation of opinions. This requires joint classi- fication in this example with respect to per capita income, age of homemaker, presence or absence of children in the home. This simple example leads to a description of the over-all procedure and the rationale of the method used in this study and a statement as to what may be expected in the way of a summary and the conclusions reached. The investigation proceeded with the following steps: 1. A list of variables was made representing an a priori judgment of those factors which are important in determining which foods are purchased, by whom, and their importance with respect 10 to the budget. In addition variables were added to describe the more subtle psychological characteristics of the homemaker. This original list contained 35 variables. 2. The apriori variables were paired in all possible com- binations and tested for significant relationships. 3. Those variables showing a high proportion of significant relationships to other variables were selected, a few new variables were formed as a result of questions raised by the tests in step 2, and a second set of 19 variables was investigated with respect to specific answers to the questionnaire exploring the role-self conflict. The newly formed variables were also paired in all possible com- binations with the apriori list. 4. Noting the socio-economic variables which tended to display a high degree of interrelationship, subgroups were formed which were homogeneous with respect to these highly interrelated variables. Specific comparisons were made within these subgroups to detect the relationships predicted by the hypotheses formed from the thesis. This step is the process of sorting out those subgroups the combination of which could obscure relationships in the manner described above. This step represents a device by means of which those subgroups responsible for specific effects can be isolated. This is the method of controlling variables in order to avoid the effects of intercorrelation. It is tacitly assumed that if a subgroup is formed within the sample 11 that is homogeneous with respect to a number of the socio-econbmic variables, then in response to the pressures of similar environmental circumstance, a number of other attitudes and characteristics will be common to the subgroup. The results and conclusions will be of four kinds. First, in the comparison of all possible pairs of the apriori variables with respect to the sample as a whole, both those relationships significant at the . 10 or less level and those showing no significance at the . 32 or less level are reported. Secondly, differences between value classes with respect to specific answers to the role-self questionnaire are noted. Third, there is the investigation of the decision energy concept and statements are made concerning the evidence for support or contradiction. As an extension to this energy concept, some notion as to the causal relationships will be advanced. Among those foods which are investigated in detail are dairy products, fruits, and eggs, all important Michigan farm products. The replies to the role-self questionnaire tend to sketch verbal profiles of the families in the low and high expenditure classes for these commodities. Even the thiamin evaluation has specific application to Michigan agriculture when it is realized that among the better sources for this nutrient are wheat germ, dairy products, and dried beans. Finally, a word will be said about the method used, its 12 advantages and disadvantages, its use in this study, and suggestions concerning future use of consumer panel data. DEFINITIONS AND DERIVATIONS OF THE VARIABLES The definitions of the terms used in this study are as follows: 1. The Sample; The sample is a group of Lansing, Michigan families who: (a) Reported food expenditures each week for the calendar year 1958 as members of the lvlichigan State University Consumer Panel. The diary kept by these families is included in the Appendix. The panel members were paid to keep these diaries under a system that included bonus payments for prompt mailing, thus minimizing recall effects. The panel sample was designed to be representative of Lansing area families. (b) Answered the Third Special Panel Questionnaire (see Appendix). The questionnaires were administered by mail after the termination of panel activities, January 1,1959. 2. Family or Household; A group of people living together with a common food budget. 3. Homemaker: The respondent in the family to the Third Special Panel Questionnaire. It is assumed that the respondent is a female adult except in the case of one person families. 13 4. Class: This term is used to designate a group of families of homogeneous value with respect to one or more variables used in this investigation. 5. Variable: A characteristic describing a family. It is bivalued (high-low), (present-absent, with respect to presence of specific age groups of children in the family). Each family was designated as being in a high or low value class with respect to each variable and each variable was examined with respect to differences between value classes in the specific phrases and adjectives used in reply to the role-self questionnaire. The description of the variables will include the method for deriving them and the value used to designate the high or low value classes. The number of people in the family, used in deriving the per capita expenditure figure and the size of family, was computed by dividing the sum of the number of people reported each week in the panel diary, by 52 weeks. It is, therefore, the average number of people in the household during 1958. The general method for determining the value used to designate the value classes was to calculate the arithmetic mean and use it as the lower bound for the high value class. This procedure is not followed in all cases. Where the data is limited such as that for per capita expenditures for prepared dinners (T. V. dinners), an 14 arbitrary division point was selected which avoided dividing the sample at the level of expenditure showing maximum frequency. In this case there was a large frequency at a very low level of expenditure but a number of families who had expenditure levels through a wide range of higher values. The arithmetic mean fell in the narrow range of high frequency at low expenditures. It was decided to treat this "cluster” of families at the low expenditure level as a group and divide the variable into its value classes somewhat. above the mean. The following describes the variables used in both steps 1 and 3 of the study: Socio-Economic Variables (a) Per capita income: Per capita income after federal income taxes in 1958; high, $2200 or more per year. (b) Per capita expenditure for food in 1958; high $287. 20 or more. (c) Age of homemaker: High, 50 years or more. This is not the mean of the sample. This point was selected to maximize inter- correlation with the presence of young children in the household. ((1) Education of homemaker: High, 12 years or more. This is not the sample mean. It is a concession to what appears as a natural bound. (e) Per capita meals at home; High, 1061 meals or more. This variable was included to detect possible false inference due to 16 (k) Number of different items included in 50 percent of the total food bill: High, 16 or more. This variable was derived by ordering the number of different items purchased in descending order of size of expenditure for the item and then counting the number of items necessary to accumulate 50 percent of the total expenditure for food. Food QuantitLand Expenditure Variables (1) Per adult expenditure for both ground and soluble coffee: High $16. 21 or more. (m) Per capita expenditure for refined sugars: High, $7. 07 or more. This group does not include molasses or honey but does include candy and confections. (n) Per capita dozens of eggs purchased: High, 24. 2 dozen or more. (0) Per capita expenditure for prepared dinners (hereinafter called T.V. dinners): High, $5.01 or more (not the mean). (p) Per capita expenditure for vitamin supplements: High, $4. 97 or more. This dividing value is above the mean due to a high frequency within a narrow range of low values which included the mean. (q) Per capita expenditure for dairy products: High, $14. 87 or more. (r) Per capita expenditure for fats and oils: High, $12. 23 or more. (5) Per capita expenditure for fruits: High, $24. 15 or more. J. . Illlli. 15 unreported food consumption arising from meals eaten away from home. (f) Presence of children age 0—9. (g) Presence of children age 10-19. Shoppirg Behavior Variables (h) Percent of total different items that were purchased only once during 1958: High, 37 percent or more. (i) Number of different items per person in family purchased during 1958. A preliminary scatter indicated that the relationship between size of family and number of different items purchased was curvilinear. This is logical since the impact of additional children in the family, after the first one, would not be expected to add to the demands for variety to the extent of the first child. In view of this, a separate mean for the number of different items purchased was computed for each size of family category. The size of family was noted for each family and the appropriate mean used to classify it as being a purchaser of a high or low number of different items during the year. The means used are as follows: No. in Family Mean No. in Family Mean (3: 1/2) (i 1/ 2) l 185 items 4 264 items 2 219 items 5 or more 264 items 3 260 items (j) Percent of total income spent for food; High, 18. 84 percent 01' more. 16 (k) Number of different items included in 50 percent of the total food bill: High, 16 or more. This variable was derived by ordering the number of different items purchased in descending order of size of expenditure for the item and then counting the number of items necessary to accumulate 50 percent of the total expenditure for food. Food QuantitJLand Expenditure Variables (1) Per adult expenditure for both ground and soluble coffee: High $16. 21 or more. (m) Per capita expenditure for refined sugars: High, $7. 07 or more. This group does not include molasses or honey but does include candy and confections. (n) Per capita dozens of eggs purchased: High, 24. 2 dozen or more. (0) Per capita expenditure for prepared dinners (hereinafter called T. V. dinners): High, $5. 01 or more (not the mean). (p) Per capita expenditure for vitamin supplements: High, $4. 97 or more. This dividing value is above the mean due to a high frequency within a narrow range of low values which included the mean. (q) Per capita expenditure for dairy products: High, $14. 87 or more. (r) Per capita expenditure for fats and oils: High, $12.23 or more. (5) Per capita expenditure for fruits: High, $24.15 or more. 17 (t) Per capita expenditure for vegetables: High, $26.19 or more. (u) Per capita expenditure for meats: High, $94. 78 or more. (v) Per capita expenditure for bakery and cereal products: High, $37. 79 or more. (w) Per capita milligrams of thiamin in food purchased: High, 1. 24 mg. per day. Estimates of the mg. of thiamin per lb. in each food item were made and multiplied by the number of lbs. of each food purchased by each family. The mean above is approximately the minimum daily requirement for the active male adult. Requirements are a function of body size, activity, and caloric intake. Psychological Variable s Source: Third Special Panel Questionnaire (the role-self questionnaire referred to in this study in parts II and III of the Third Special Panel Questionnaire). (aa) Conflict score: High, 10 conflicts or more. In the first half of the role-self questionnaire the questions essentially asked, "What kind of home should a homemaker run and what activities should she consider important ?” In the second half the questions asked, "What kind of home do 133 run and what activities do 1211 consider to be important ?” The replies in each case were selected from a list of ten adjectives or phrases that followed each of a total of 16 questions (see Appendix p. 108). The respondent was requested to select the four most appropriate adjectives or phrases. The questions and 18 answers in each of the two parts of the role-self questionnaire were matched although the adjectives and phrases were not in the same order for the matched questions. In deriving this variable, each instance in which the adjective used to describe the role (first section) was not used to describe the self (second section) was scored as a conflict. The total of conflicts of this nature was summed for each respondent and assigned as the conflict score of the respondentgs family. The arithmetic mean of these scores was used as the high-low dividing value. This conflict score as developed here is the same as that . 13 . . . used by Trier. In his study he evaluated this measure in terms of a personality inventory scale applied jointly to another sample of Lansing l4 , homemakers. Trier concluded that: The greater the amount of conflict in role perceptions, it was found, the poorer the personality picture. Wives with more role conflicts were found to be more pessimistic, more emotional, more disorganized, and less self—confident. 15 In developing this measure, Trier, in his thesis, cites other studies which associate high role conflict with neurotic tendencies. If it is asserted that personality traits associated with high conflict scores 3Trier, Howard E. ”Sociological Variables, Personality Traits, and Buying Attitudes Related to Role Perception and Conflicts: Among 242 Michigan Housewives, ” Unpublished Ph. D. thesis, Michigan State University, 1959. 1 4This sample of 242 homemakers was drawn from a larger random sample of 1200 Lansing, Michigan, housewives. The homemakers were surveyed by personal interview in the summer of 1958. 1 5Trier, op. cit., pp. 120-21. 19 represent variations in energy that might be used in decision-making, then it should be possible to support hypotheses associating conflict scores with shopping attitudes or behavior that imply active decision- making. To indicate more clearly what is meant by "role" and "self" the following are quoted: From Mowrer, Role refers to that pattern or coordination of attitudes and habits through which the individual secures and maintains status in a particular group. Since an individual belongs to many groups, in each of which he plays a different role, his ~ personality is made up of many roles. I From Snygg and Combs, in discussing the self concept, What a person does and how he behaves is determined by the concept that he has of himself and his abilities. It (the self) gives consistency to his behavior. It is the only frame of reference that the individual possesses. 17 (bb) Social score: High, 7 or more. This characteristic is subjective in nature and derived internally from replies to the role-self questionnaire. It was noticed in examining the replies to the questions that 5 or 6 adjectives were used by all of the homemakers for a majority of the answers. The use of atypical answers was assumed to be of possible import and these were isolated. Upon isolation it was noticed that the atypical answers or phrases were generally associated with 16 . . . . Mowrer, Harriet. Personality Adjustment and Domestic Discord. New York: American Book Co. , 1935, as quoted in Trier thesis. 7Snygg, D. and Combs, A. W. Individual Behavior. New York: Harper 8: Bros. , 1949, as quoted in Trier thesis. 20 interests and activities outside of the home, or bringing into the home persons from outside the family. Examples of such replies are as follows: (a) I run a fashionable home. (b) Entertainifl friends and associates should be an important activity of the homemaker. (c) Helpingmy husband advance in his job is one of my important duties. (d) I spend much of my time participating in civic activities. The adjectives that did not conform to this vein of thought were dropped, the number of remaining atypical adjectives or phrases used were counted for each respondent, and the total frequency of use of atypical adjectives and phrases was assigned to the family as the social score. Variables Indicating Attitudes toward Shopping: Source Part I of Third Special Panel Questionnaire These questions explore the attitude of the homemaker toward shopping for food. The questions are oriented toward the 10 shopping decision components indicated below. The answers to the questions from which the scores are derived are a series of statements ranging in meaning from a strongly positive attitude to a strongly negative one. The replies were coded by means of the numbers 1 to 5 as the replies progressed from strongly negative to strongly positive. The score for each family with respect to each variable was computed by summing the 5 code numbers representing replies. A high sum inc-u..." .tIIDII‘r!‘ eel) tli . 21 indicated a preponderance of positive attitudes, a low sum indicated relatively negative replies. The mean of the scores for all families for each variable was computed and used to establish the high and low value classes. The attitudes and dividing scores are as follows: (cc) Price oriented: High, 17 or more. (dd) Quality oriented: High, 16 or more. (ee) Convenience oriented: High, 9 or more. (ff) Pre-planning oriented: High, 16 or more. (gg) Information-seeking oriented: High, 14 or more. (hh) Calculation oriented: High, 15 or more. (i i) Other (people) oriented: High, 14 or more. (jj) Wife oriented: High, 11 or more. (kk) Husband oriented: High, 5 or more. (11) Parent oriented: High, 12 or more. The general import of these questions and the associated variables is most efficiently drawn by reading the questions. A list of these questions organized with respect to the intended orientation is included in the appendix for this purpose. In addition, in front of each question a percentage figure is given indicating the frequency of replies to each as found in the replies of another sample of Lansing families (cf. footnote 14, p. 18). This is included to indicate the reaction to these questions by a random sample drawn from the same population ' as the sample used in this study. As an example of the derivation of 22 the above shopping attitudes; highly price oriented shopping attitudes were derived from statements to the effect that the homemaker is highly aware of the prices of food items and is sensitive to price change. Highly convenience oriented shopping attitudes are derived from a high incidence of affirmative or positive answers to questions asking whether or not time and effort saving in meal preparation is a criterion for the purchase of food. TESTS FOR SIGNIFICANCE The use of a bivalued classification system leads to the use of the binomial probability distribution as the appropriate statistical model on which to base the decision rules for acceptance or rejection of hypotheses and the tests for significance. The description of a family with respect to 2 characteristics each with two values, yields a sample space equivalent to that of El trials with two outcomes. This is the multivariate binomial distribution and the probability associated with a given description or combination of values is the product of the relevant marginal probabilities, if the characteristics are independent of each other. To illustrate the functioning of this type of model in its simplest form, and in a form which is used in the analysis of subgroups within the sample, the con- struction and testing of 2 x 2 tables follows: If a population of 100 families is considered jointly with respect to per capita income and per capita expenditure for food and 23 only two-values are admitted for each of these characteristics the possible combinations of these two variables may be represented by the cells in the table below. (By using 100 families the frequencies which appear as the lowest figure in the cell, become the percent of the sample possessing the joint characteristics which the cell represents.) Per Capita Income low high '0 f - o L: low, low high, low (,3 ,4 low p=.20 p=.30 50% p=.50 E. 8 40 10 w e “3. a 3: lg low, high high, high a high p=.20 p=.30 50% p=.50 >< » O 50 £11 40% 60% Marginal p = .40 p = .60 Probabilities The above table demonstrates a significant relationship between low per capita income and low per capita expenditures for food. Since a sample of 100 families has been used, the values of p in each cell multiplied by 100 are the expected frequencies if the two variables are completely unrelated. The value of p is the probability that a family in the sample will possess the characteristics indicated by the cell specifications. Note that with the size of the sample and the marginal probabilities given, 24 a single entry determines the whole table. No further information is needed to complete it. i The investigation of all possible pairs of the original 35 apr'ior'i variables led to potential formation of 595 tables of dimension 2 x 2. To reduce the work involved in computing all four cells for each and the chi-square tests for each of these tables, the fact that the table, with given marginal probabilities, is completely determined by a single entry, was used. As the total frequency of such a table increases,the binomial distribution lying behind its construction approaches the normal distribution. Hence in the analysis of the sample as a whole, only the low, low cell was observed, its expected frequency computed (p x No. in sample) and the deviation of frequency of the observed cell measured according to the following formula: 5 _ Z : —n——;—§ Where: Z :2 Number of standard deviations ,1, (mm) n :: number in the sample In the preceding table: p 2: joint probability associated with n = 100 tip 2 20 the (low, low) cell p=.20 Z_40-20_5 q::(l-p) q = . 80 (16)1: Z np :: expected frequency in the (low, low) cell Sn : 40 Sn 3: observed frequency in the (low, low) cell Consulting a table of the normal distribution p(Z) < . 0001 25 Decision rule: The hypothesis that the variables are independent is rejected if: Z > 1. 645 at the . 10 level of significance Z > 1. 960 at the . 05 level of significance Z > 2. 575 at the . 01 level of significance Unless otherwise specified the term significant will mean significance at the .10 or less level. In the analysis of the subgroups within the total sample, since the numbers of families under analysis was smaller, the chi- square test was used. Chi-square was computed by means of the following formula containing the l_l_ correction: 2 _ (lad-bc' -1/2n)2n ' (a+b)(a+c)(b+d)(c+d) if the cells are : Kg Decision rule: Reject the hypothesis that the variables are independent if: (one degree of freedom) 2 x > 2. 706 at the .10 level of significance 2 X > 3. 841 at the .05 level of significance 2 X > 6. 635 at the . 01 level of significance The term significant will mean significant at the . 10 level unless otherwise specified. Asterisks (*) will be used to denote levels of significance of lesser numerical value than . 10 and are uniform for all analyses in this study. The use of these will be demonstrated in subsequent footnote 5 . 26 ASSUMPTIONS IN THE STUDY Certain limitations in the data place restrictions on the inference which may be made from observed relationships. The following observations and assumptions indicate where the bounds to such inferences lie: 1. There is the question of why expenditure rather than quantity datawe-re used for the food consumption variables. The first answer is simply that with quantity data for diverse foods the measure becomes meaningless. Secondly, and more to the point, shopping implies a decision related to allocation of money income, hence the variables are indicated in monetary terms. 2. The consumption data are a total of expenditures for a whole family reporting to the M. S. U. Consumer Panel but only the homemaker answered the Third Special Panel Questionnaire used to assess role-self characteristics and attitudes toward shopping. The inference that food purchased is associated with psychological scores through a relationship in common with the nutritional content of the food, requires the assumption that the homemaker's diet is either similar to the diet of the family as a whole or deviates from the diet of the balance of the family in a consistent manner within a given socio- economic subgroup of the sample. The inclusion of such an assumption is not as damaging as it might seem. Much of human behavior is directed toward or is in reaction to the behavior of other people and 27 hence the wife would not behave as if insulated from the balance of the family. Since the immediate concern is the identification of differences rather than causality, all that is necessary is the reflection of differences in the families' behavior and attitudes in the homemakers' responses to the questionnaire. 3. Another assumption is concerned with the fact that the data were gathered for the year 1958 while the questionnaire from which the psychological variables were derived was mailed and returned in January, 1959. It is assumed that the properties of the variables derived from the expenditure data and the role-self and attitude questionnaire are sufficiently stable to establish continuity between the two sets of data. 4. It is assumed that the verbalizations in the questionnaire are meaningful such that answers by an individual would be consistent if repeated without memory, and that the interpretation of a particular question within homogeneous subgroups is similar. 28 CHAPTER II RESULTS OF THE ANALYSIS OF THE FULL SAMPLE As outlined in the previous introductory statements, the search for the kinds of relationships indicated by the initial questions which define the problem area proceeded in two steps. The first of these was the examination of all possible pairs of variables, including the ones formed after the apriori group, with respect to the full sample. In addition there was a simultaneous examination of the differences in specific replies to the role-self questionnaire between high and low value classes for each variable. There are two purposes for conducting this first phase of the investigation. The first is to find and report those relationships which are relevant to the thesis and true for the sample as a whole. The second purpose is to determine those demographic relationships that are highly interrelated and which will formthe groups of variables used in constructing meaningful subgroups, the analysis of which is reported in the next chapter. As will be seen, these variables in general define specific groups of families with respect to temporal position in the life cycle of family formation, development, and maturation. The exact form and order of reporting is as follows: 29 1. Each variable is considered in turn. Each relationship except for a few highly interrelated shopping attitude variables is reported twice, once for each variable involved in the pairing. 2. The second result reported is the differences in replies to the role-self questionnaire between the value classes for the variable. Significance was established by the application of the t test. Those relationships reported are significant at the . 05 or less level. More emphasis means that more than the expected number of homemakers used a given phrase or adjective. Less emphasis means that fewer than the expected number of homemakers used the phrase or adjective. Considerable caution should be used in interpreting these differences in emphasis. Where an adjective or phrase is given less emphasis it does not mean that there is a positive denial of the appropriateness of this adjective or phrase. It means that other replies are more appropriate. More explicitly, an example of an equivalent composite statement would be, "It may be that I run a quiet home but I think the words comfortable, modern, fashionable, efficient are more appro- priate in describing my home. " 3. The second group of relationships reported for each variable is the existence, or lack of existence of, significant association between the value classes of all possible pairs of variables. Since the results are derived from 2 x 2 tables the opposite of relationships resulting from examining a low (present) class will be 30 true for the high (absent) value class for the same variable. Examples (a) Adjectives or phrases in the role-self questionnaire which are given more emphasis by a low value class are given less emphasis by the high value class. (b) If the low value class of variable A is significantly related to the low value class of variable B, then the high variable class for A will be significantly related to the high value class of B. (c) If the low value class of variable A is significantly related to the high value class of variable B, then the high value class of A is significantly related to the low value class of B. PER CAPITA INCOME The low per capita income class contains families of two typical sizes. The following table shows the frequency distribution for both value classes of this variable by size of family. Average Family Number Families Number Families Size High Low 0 50-1. 49 16 10 1. 50-2. 49 33 42 2. 50-3. 49 14 19 3. 50-4. 49 3 37 4. 50-5. 49 0 24 5. 50 or more 0 11 It is seen that in the low value class that two and four person families predominate. In addition, the two person low income 31 family group contains both elderly couples, presumably on retirement incomes, and younger couples who normally have their peak income rates ahead of them. The tendency of these subgroups to present conflicting relationships, will, to some degree tend to obscure some of the unique aspects of the low per capita income class. The analysis of these subgroups in isolation is the subject matter of the next chapter. The other remark necessary with reference to the discussion of income classes, is the disparity of economic restriction on families of different size within the same per capita income class. As an example, an income of $4400 per year for two persons does not usually result in the same standard of living as $11, 000 per year for two adults and three children. Keeping these characteristics of this income aggregation in mind, the results of the tabular analysis were as follows: In replying to the role-self questionnaire, low income families gave more emphasis to: 1 Role 9 Self20 Disciplining the children Disciplining the children Skill at baking Skill at canning Spending time keeping the Enjoyment of raising children house clean. Enjoyment of children Spending time with children. 18 More emphasis by low income class means less emphasis by high income class. Less emphasis by low income class implies more emphasis by high income class. Role—self reporting for all variables is at the . 05 level of significance. 32 . . . . 21 Low income families gave less emphaSis to: Role _;. Self Helping husband's career Being intelligent Being intelligent Entertaining friends Being informed on new ideas Skill at baking Spending time keeping own Being informed on new ideas appearance trim Spending time earning additional family income Keeping own appearance trim The replies to the role-self questionnaire indicate that the low income class is compositely occupied with child rearing. Equating skills at canning and baking, the only conflicts appearing in the role concept vs. the self is over the matter of spending time cleaning house and helping husband's career, since these appear in the rolebut not the self concept. Relating per capita income to the other variables resulted in the following: 9In interpreting "role" the central idea is the phrase "A homemaker should . . . . " This is an assessment of what values the homemaker considers appropriate to the role of homemaker. 0 l Interpreting "self" is best understood by using these phrases as direct description by the homemaker of herself, classified as ”more" or ”less" appropriate. More emphasis by low income class means less emphasis by high income class. Less emphasis by low income class implies more emphasis by high income class. Role-self reporting for all variables is at the . 05 level of significance. 33 Low per capita income families were significantly associated with: Low per capita expenditure for food* Presence of children age O-9*, 10-19 Low per capita number of different items purchased Low per capita expenditure for fruits, vegetables, meat* Low per capita mg. thiamin in food purchased High percent of income spent for food. * Neither value class for per capita income was associated with the following at the . 32 or less level of significance: Expenditure for: coffee, vitamin supplements, T.V. dinners, dairy products, bakery goods Dozens of eggs purchased Conflict score Shopping attitudes oriented toward: price, quality, convenience, wife, husband. The results indicate that per capita income is related to expenditures for some, but not all, classes of foods. Among those listed as not related are two principal sources of protein, dairy products and eggs. The relationship of low per capita income to high percent of income spent for food will be used in the subgroup analysis to follow . 01 level of significance. 22Those variables not listed as significantly related, or unrelated at the . 32 or less level, fall between . 10 and . 32 levels of significance. 34 in which percent of income spent for food is used as an alternative indicator of economic pressure on the food budget. The observation that price, quality and convenience oriented shopping attitudes are unrelated to per capita income will be seen to be the result of sub- groups holding conflicting attitudes, thus cancelling relationships. PER CAPITA EXPENDITURE FOR FOOD The frequency distribution with respect to family size follows: Family Size Number of Families~« Number of Families High Low 0. 50-1. 49 17 9 1. 50-2. 49 46 29 2. 50-3. 49 27 6 3. 50—4. 49 25 15 4. 50-5. 49 5 19 5. 50 or more 1 9 Examining replies to the role-self questionnaire indicated 23 that the class, low per capita expenditure for food, gave more emphasis to: Role (A homemaker should) Self (I am) Disciplining children Disciplining children* Being planful Enjoyment of raising children* Spending time with children. 23. 05 level of significance. . 01 level of significance. 35 The class low per capita expenditure for food gave less , 24 emphaSis to: Role (A homemaker should) Self (I am) Enjoyment of purchasing food. Being calm Keeping own appearance trim. The above relationships imply that the high value class for per capita expenditure for food gives increased emphasis to enjoyment of shopping for food in the role concept but not in the self concept. The increased emphasis given to being planful in the self concept of the low value class, coupled with a lack of relationship between the low value class and pre-planning oriented shopping attitudes, indicates that planful assumes a meaning broader than budgeting food purchases. This same apparent contradiction appears with respect to the variables, age of homemaker, and per capita number of different items purchased. Relating per capita expenditure for food to the other variables, low per capita expenditure for food was significantly associated with: Low per capita income* Low age of homemaker Low number of different items purchased Low per capita expenditure for: sugar, * dairy products, fats and oils, * fruits, * vegetables, * meats* 24. 05 level of significance. *. 01 level of significance. 36 Low per capita dozen of eggs purchased Low per capita mg. of thiamin in food purchased* High calculation oriented shopping attitudes* High percent of income spent for food. * Neither value class for per capita expenditure for food was related to the following at the . 32 or less level of significance: Education of homemaker Per capita meals served at home Expenditure for coffee Conflict score Social score Shopping attitudes oriented toward: convenience, pro-planning, information-seeking, other (people), husband, parents. All of the food expenditure classes were related to per capita expenditure for food except coffee. Although low per capita expenditure for food is related to highly calculation oriented shopping attitudes there was no significant association with attitudes oriented toward pre-planning or information-seeking. This, in the light of the above self concept of being planful seems to indicate that the food shopping decision is made largely in the store. * . 01 level of significance. 37 AGE OF HOME MAKER The frequency distribution by family size follows: , , Number of Families Number of Families Family Size , High Low 0. 50—1. 49 22 4 1. 50-2. 49 61 14 2. 50-3. 49 12 21 3. 50-4. 49 3 37 4. 50-5. 49 4 20 5. 50 or more 0 11 In replying to the role-self questionnaire the younger homemakers gave more emphasis to: Role Self Disciplining the children Disciplining the children Being planful Being planful Being progressive Running a relaxed home Being a sexual companion Being a companion to children to husband Enjoyment of raising children Being a companion to children Enjoying children Enjoyment of raising children Spending time with children. Enjoying children Spending time with children. The younger homemakers gave less emphasis to: 25 . . Differences in replies to role-self questionnaire significant at the . 05 level. 38 Role Self Mending clothes Running a quiet home Kindness Being self-controlled Handling family finances Entertaining friends Skill at baking Handling family finances Skill at canning Being calm Enjoying shopping for food Enjoying shopping for food Sewing. Spending time shopping for bargains Spending time planning meals. The class of homemakers under 50 years of age is a distinct one in terms of differences in adjectives and phrases used to describe role and self concepts. Ten of the fifteen phrases given greater emphasis mention children directly. Being planful is given greater emphasis by the younger homemakers in both the role and self. They give lesser emphasis to handling family finances and enjoyment of shopping for food in the role concept, and handling family finances, enjoyment of buying food, spending time shopping for bargains, and planning meals in the self concept. This supports an assertion that planfulness involves a concept of strategy other than short term allocation of money and time with respect to food and finances. It was also observed that a cluster of activities, sewing, mending clothes, skill at baking, and skill at canning are given less emphasis by the younger homemakers. This, taken together with the lack of significant relationship between age of 39 homemaker and price, calculation, or convenience oriented shopping attitudes, would indicate that the younger homemakers are willing to purchase such household services. Finally, the younger homemaker does not see herself to be self-controlled, or calm, when compared to the concepts held by the older homemaker. With respect to the other variables, the class of home- makers less than 50 years of age was significantly associated with: Low per capita expenditure for food Low number of meals eaten at home Children present, age 0-9,* 10—19* Low percent of total items purchased only once Low per capita expenditure for: sugar, fats and oils, fruits, vegetables, bakery goods’l‘ Low per capita dozen of eggs purchased Low per capita mg. of thiamin in food purchased* High education of homemaker. * Neither value class for age of homemaker was related to the following at the . 32 or less level of significance: Expenditures for T. V. dinners, dairy products, meat Shopping attitudes oriented toward: price, quality, convenience, pre-planning, information-seeking, calculation, other (people), husband, parents. >5: . 01 level of significance. 40 The pattern of expenditure for food by the younger home- makers resembles that for families with low per capita income, high percent of income spent for food, and children age 0-9 in the household. It is along these lines that the subgroups in Chapter 3 are formed. EDUCATION OF HOMEMAKER This variable showed little significant relationship to the other variables in the investigation and was not examined with respect to the role-self questionnaire. In general, those persons with less than completed high school educations, were elderly people in the low per capita income class and this group was numerically small. Where a significant relationship does occur, it is noted in the discussion of the other variable . NUMBER OF MEALS SERVED AT HOME This variable was also dropped afterapriori consideration and was not related to the role—self questionnaire. In relating this variable to the other variables it was found that those families serving less than 1061 meals at home during 1958 were significantly associated with: Low percent of income spent for all food Low age of homemaker Presence of children age 10-19 Low expenditure for bakery goods Low per capita mg. of thiamin in the food purchased 41 High education of homemaker Neither value class for this variable was related at the . 32 or less level of significance to any of the remaining variables except those listed as significant above and: Per capita expenditures for sugar Per capita dozen of eggs purchased Shopping attitudes oriented toward wife, parents. The general lack of relationships for this variable leaves only a hint as to its nature. The dividing value is at an average of 2. 85 meals per day, and the low value class aligns itself with the group of homemakers under 50 years of age with children 10-19 in the household. The low value class for this variable is also associated with low per capita expenditure for bakery goods. Since the low value class is formed at values that are considerably above two meals per day, and observing the age of the children, it could well be that the variation in number of meals served at home reflects an occasional meal skipped on school days. PRESENCE OR ABSENCE OF CHILDREN AGE 0-9 The frequency distribution for this variable with respect to family size was: 42 Size of Family Number of Families - Number of Families- Children Absent Children Present 0. 50- 1. 49 26 0 1. 50-2. 49 75 O 2. 50-3. 49 24 9 3. 50-4. 49 14 26 4. 50-5. 49 4 20 5. 50 or more 0 11 Examination of the replies to the role-self questionnaire showed that the class of homemakers with children age 0-9 in the house- hold give more emphasis to (. 05 level of significance): Role Disciplining the children Being planful Being a companion to the children Enjoyment of raising children Spending time with children Self Disciplining the children Being a companion to the children Enjoyment of raising children Enjoying children Spending time with children The homemaker with children age 0-9 in the household gave less emphasis to (. 05 level of significance): Role Mending clothes Enjoyment of cooking Kindness Spending time planning meals Spending time keeping own appearance trim Self Running a cheerful home Running an efficient home Importance of meal planning Being self-controlled Entertaining 43 Role Self Enjoyment of buying food Being calm Enjoyment of cooking Spending time shopping for bargains Spending time earning extra family income Spending time planning meals Spending time keeping own appearance trim. The principal features of these observations are the pre- dominance of the mention of children in the phrases given increased emphasis, and the predominance of the mention of time and household chores in the phrases given lesser emphasis. These homemakers do not see themselves as being calm or self-controlled, nor did they tend to use the words cheerful or efficient to describe the way they run their homes. Note that this group was also significantly associated with low conflict scores below. In relating this variable to other variables, those families with children age 0-9 in the household were significantly associated with the following: Low per capita income* Low per capita expenditure for food* Low age of homemaker* . 01 level of significance. 44 Presence of children age 10-19* Low percent of different items purchased only once Low per capita expenditure for: sugar, fats and oils, fruit, * vegetables, meat, bakery goods Low per capita mg. of thiamin in the food purchased Low conflict score Low social score High education of homemaker* Shopping attitudes highly oriented toward: preplanning, husband. * Neither value class for this variable was related to the following variables at the . 32 or less level of significance: Number of different items included in 50 percent of the food bill Per capita expenditure for: T. V. dinners, coffee, dairy products Per capita dozen eggs purchased Shopping attitudes oriented toward: price, quality, convenience, information-seeking, calculation, other (people). Here as in the analysis of the class, low age of homemaker, being planful is given more emphasis by the homemakers in the low value group. With this variable, presence of children, it does not appear in the self concept. On the other hand, unlike the low age of homemaker class, the homemakers with young children show pre- planning oriented shopping attitudes. Information seeking and calculation oriented attitudes are not significantly associated with these homemakers. . 01 level of significance. 45 Again the term planful seems to have somewhat disjoint meaning from meal planning or budgeting. The husband oriented shopping attitude is consistent the the role concept giving a lesser emphasis to enjoyment of shopping. The relationship to eggs, dairy products, and meat are the same as for the low per capita income class. The obvious interrelationship among this variable, age of homemaker, per capita income and per capita expenditure for food, is the basis for subgroup formation. PRESENCE OR ABSENCE OF CHILDREN AGE 10-19 This variable was not examined with respect to the role- self questionnaire. The relationship to the other variables was such that the families with children age 10-19 in the household were signifi- cantly associated with: Low per capita income Low age of homemaker* Low number of meals served at home Presence of children 0-9 Low percent of different items purchased only once Low per capita expenditure for fruits High per adult expenditure for coffee. Neither value class for this variable was related to the following variables at the . 32 or less level of significance: . 01 level of significance. 46 Per capita number of different items purchased Percent of income spent for food Number of different items included in 50 percent of the food bill Per capita expenditure for: T. V. dinners, dairy products, vegetables, meat Per capita dozen eggs purchased Conflict score Social score. The unique features are the previously noted relationship to meals served at home, and the relationship to expenditures for coffee. The latter indicates the effect of the children starting to participate in the drinking of this beverage. The presence of the children age 10-19 unlike the variable presence of children 0-9, is not significantly related to most per capita food expenditures. PERCENT OF TOTAL NUMBER OF DIFFERENT FOOD ITEMS THAT WERE PURCHASED ONLY ONCE IN 1958 This variable was formed to identify families in terms of a workable measure of the pattern rather than just the number of repeated purchases. In joint classification with the value classes of the variable, per capita number of different items purchased, it is possible to identify families whose purchases tend to be composed of a large number of singly purchased items. These variables, jointly with the variable per capita expenditure for food, would indicate a degree of concentration of expenditures on relatively few items. 47 This procedure is used in the subgroups to try to identify habitual purchasing activities. In using the variable, percent of different items purchased only once, singly, and with respect to the sample as a whole, one must recognize that a large percent of items purchased only once, could arise from a large number of different items purchased, sensitivity to price movements, or a tendency on the part of the homemaker to do experimental buying. The results in this study describe the performance of this variable as much as they furnish a basis for the empirical measurement of habitual shopping behavior. In reply to the role-self questionnaire those homemakers who purchased the lower percent of items only once, gave more emphasis to (. 05 level of significance): Role Self Companionship to husband Disciplining the children Being a sexual companion to husband Being ambitious Enjoyment of raising children Spending time with children The homemakers who purchased a lower percent of items only once gave less emphasis to (. 05 level of significance): 48 Role Self Mending clothes Running a quiet home Helping husband on job Being informed on new ideas Skill at baking. Spending time shopping for bargains. These replies resemble those of the younger, low per capita income, homemaker with the younger children in the household. In examining this variable with respect to the other variables it was found that the homemakers who purchased the lower percent of total items only once were associated with the following value classes: Low age of homemaker Presence of children 0-9, 10-19 Low social score High percent of income spent for food Highly price oriented shopping behavior. The value classes for this variable were unrelated to the following variables at the . 32 or less level of significance: Per capita expenditure for all foods except: sugar, coffee, dairy products All shopping attitude variables wt: price oriented Conflict score Per capita mg. of thiamin in food purchased Per capita number of different items purchased Number of items included in 50 percent of food bill. 49 The relationship of this variable to price oriented shopping attitudes tends to discourage the hypothesis that a high percentage of items purchased only once is due to price sensitivity. The relationships above indicate that the homemaker who purchases a large percent of items only once is the older homemaker with no children in the household and whose interests lie outside of the home environment. NUMBER OF DIFFERENT ITEMS PURCHASED As outlined in defining this variable the number of different items purchased was observed to be a positive curvilinear function of family size with a maximum at the four person family level. The high vs. low dividing value was the mean as determined within each size of family group. The families in which a low per capita number of different items were purchased were found to be significantly associated with replies to the role-self questionnaire giving more emphasis to (. 05 level of significance): Role Self Being planful Being planful Handling family finances Shopping for bargains Planning meals. Planning meals. 50 The homemakers in the class, low per capita different number of items purchased, gave less emphasis to (. 05 level of significance): Role Self Being intelligent Running an orderly home Being self-controlled Keeping own appearance trim. Families purchasing a low per capita number of different items were seen to be associated with role and self concepts which gave more emphasis to planning. The kinds of phrases given less emphasis were similar to those of the younger homemaker with young children. Relationships below, however, indicate that this emphasis on planning was not corroborated by association of the same value class with shopping attitudes oriented toward, price, preplanning, calculation, or information-seeking. Again planning seems to carry rather broad meaning. With respect to the other variables the class, low per capita number of different items purchased, was significantly associated with: Low per capita income Low per capita expenditure for food Low number of different items included in 50 percent of the food bill 51 Low per capita expenditure for: sugar, vegetables. * Neither value class for this variable was associated with the following variables at the . 32 or less level of significance: Age of homemaker Number of meals served at home Presence of children age 10-19 Percent of different items purchased only once Per capita expenditures for: coffee, vitamin supplements, T. V. dinners, dairy products, fruits, bakery goods Shopping attitudes oriented toward: price, convenience, pre- planning, information-seeking, calculation, husband, wife, parents Per capita dozen of eggs purchased Social score Conflict score. This variable is characterized by a general lack of relation- ship to the other variables noting particularly that it is unrelated to the percent of items purchased only once. It is significantly related to per capita income and per capita expenditure for food. PERCENT OF INCOME SPENT FOR FOOD Examination of the replies to the role-self questionnaire indicated that the homemakers who spent a low percent of income for food gave more emphasis to (. 05 level of significance): >3 . 01 level of significance. ti 52 Role Self Running a relaxed home Being thrifty Serving a variety of meals Being informed on new ideas. Enjoy buying food Keeping own appearance trim. The homemakers who spent a low percent of income for food gave less emphasis to (. 05 level of significance): Role Self Skill at baking Running a quiet home Skill at canning. Enjoyment of raising children. The phrases used in reply to the role-self questionnaire indicate that homemakers who spend a low percent of income on food are not under severe budget pressure, but they use the term thrifty to describe themselves. They also give increased emphasis to being informed on new ideas but there is a lack of relationship below to shopping attitudes oriented toward information-seeking. With respect to the other variables, the class low percent of income spent for food was associated with the following: Low number of meals served at home* High per capita income* High per capita expenditure for food . 01 level of significance. 53 High education of homemaker* High percent of items purchased only once High number of items included in 50 percent of the food bill High per capita expenditure for fruits. Neither value class for this variable was associated with the following variables at the . 32 or less level of significance: Age of homemaker Presence of children age 10-19 Per capita expenditure for foods except: fruits, vegetables, bakery goods Social score Conflict score Shopping attitudes except those oriented toward calculation. The above relationships indicate that the class, low percent of income spent for food, consists of families in the higher income brackets. The relationship to fruit expenditures is typical of the older homemaker. The association of low percent of income spent for food with low number of meals served at home, and lack of association with social score, suggests that meals not served were simply skipped, since phrases leading to a high social score had to do with activities away from the home. . 01 level of significance. 54 NUMBER OF ITEMS INCLUDED IN 50 PERCENT OF THE FOOD BILL This is the fourth variable under the general classification, shopping behavior, and designed to explore the notion of habit. In replying to the role-self questionnaire the only replies showing significant differences between the high and low value classes were the following. Homemakers from families with a low number of items included in 50 percent of the food bill gave more emphasis to (. 05 level of significance): Self: Enjoyment of raising children. The same class of homemakers gave less emphasis to (. 05 level of significance): Self: Being ambitious. With respect to the other variables, low number of items included in 50 percent of the food bill, was significantly associated with: Low number of different items purchased Low per capita expenditure for vegetables High education of homemaker High percent of income spent for food. Neither value class was associated at the . 32 or less level of significance with: Per capita expenditure for food Age of homemaker 55 Number of meals eaten at home Presence of children age 0-9, 10-19 Percent of different items purchased only once Per capita expenditure for all food classes except: sugar, dairy products, vegetables, bakery goods Per capita mg. of thiamin in food purchased Social score All shopping attitude variables. This variable showed few significant relationships and those that did appear are similar to those for number of different items purchased. PER ADULT EXPENDITURE FOR COFFEE The only phrase for which a significant difference (. 05 level) was found in replies to the role-self questionnaire by the value classes for this variable was a lesser emphasis given mending clothes in the self concept for the low value class. Neither value class for per adult expenditure for coffee, was significantly related at the . 32 or less level to any of the other variables except: Presence of children age 10—19. (Significant at . 05 level.) Age of homemaker Percent of different items purchased only once Per capita mg. of thiamin in food purchased 56 Per capita expenditure for meats Shopping attitudes oriented toward: price, wife. There is little basis for identifying a ”typical" coffee drinker. Presence of children 10-19 was related to high per adult expenditure for coffee. This was the only variable related to per capita coffee expenditure at the . 10 or less level of significance. PER CAPITA EXPENDITURE FOR SUGAR AND CONFECTIONS In replying to the role-self questionnaire, homemakers in the low value class gave more emphasis to (. 05 level of significance): Role Self Disciplining the children Disciplining the children Being calm. Enjoyment of raising children Enjoyment of socializing with the neighbors Spending time with the children. The low value class gave less emphasis to (. 05 level of significance): Role Self (No significant differences.) Being self—controlled Being calm Spending time keeping own appearance trim. r353“; 57 Role conflict appears over the term calmness, and the homemaker in the low expenditure class does not tend to use the terms calm and self-controlled in describing herself. The class, low per capita expenditure for sugar and con- fections, was significantly related to: Low per capita expenditure for food Low age of homemaker Presence of children age 0-9 Low number of different items purchased Low per capita expenditure for: fats and oils, * fruits, * vegetables, meats, * bakery goods. * Neither value class was related at the . 32 or less level of significance with: Education of homemaker Per capita expenditure for: coffee, vitamin supplements, T. V. dinners Social score All shopping attitudes except husband oriented. It will be seen in the reports on the per capita expenditures for food which follow that, in general, low per capita expenditure classes for a food item are related to low per capita income, low per capita expenditure for food, low age of homemaker, presence of children age 0-9, and the low per capita expenditure classes for the other food . 01 level of significance. 58 classes. The per capita expenditure for sugar and confections follows this pattern. PER CAPITA DOZENS OF EGGS PURCHASED In replying to the role-self questionnaire, homemakers from families who purchased a high number of dozens of eggs per capita gave more emphasis to (. 05 level of significance): Self: Handling family finances, decorating the home. This type of reply supports the prior assertion that high per capita egg purchases associates with older homemakers, in the higher income brackets. This is further supported in the following observations. The class, high number of dozens of eggs purchased, was significantly associated with: High per capita expenditure for food High age of homemaker High per capita expenditure for: fruits, meat High per capita mg. of thiamin in food purchased. * Neither value class was associated with the other variables at the . 32 or less level of significance except for those listed above and * i the following: 6 . . . 2 Where the list of unrelated variables is extenSive the exceptions to the condition, unrelated at the . 32 or less level, will be reported. These include those significant at the . 10 or less level and those showing significance at levels between . 32 and . 10. . 01 level of significance. 59 Number of meals eaten at home Per capita expenditure for: sugar, dairy products, fats and oils, fruits, vegetables, meats Shopping attitudes oriented toward: calculation, convenience. PER CAPITA EXPENDITURES FOR VITAMIN AND MINERAL SUPPLEMENTS In replying to the role-self questionnaire the class, low per capita expenditures for vitamin supplements, gave more emphasis to (. 05 level of significance): Role Self (No significant differences.) Disciplining the children Enjoyment of raising children Spending time with the children Serving a variety at meals Enjoyment of socializing with neighbors. The same low value class gave less emphasis to (. 05 level of significance): Role Self Mending clothes. Running a cheerful home Keeping social responsibilities . The above pattern is that of a young homemaker with children. 60 The class, low per capita expenditure for vitamin supple- ments, was significantly associated with: Low per capita mg. of thiamin in food purchased Low per capita expenditures for dairy products. The value classes for this variable were unrelated to the other variables Ex_c_e_p_t for those listed above and the following, at the . 32 or less level of significance: Per capita expenditure for food Age of homemaker Presence of children age 0-9, 10-19 Per capita expenditure for fruits Social score Shopping attitudes oriented toward price. Note that the families characterized by high per capita mg. of thiamin in the food purchased are also characterized by high per capita expenditures for vitamin and mineral supplements. PER CAPITA EXPENDITURES FOR T. V. DINNERS The value classes for this variable were examined with respect to the other variables only, and no significant relationships were found. The value classes were unrelated to the other variables at the . 32 or less level except for shopping attitudes oriented toward price. 61 PER CAPITA EXPENDITURES FOR DAIRY PRODUCTS In replying to the role—self questionnaire the homemakers in the class, low per capita expenditure for dairy products, gave more emphasis to (. 05 level of significance): Role Self Skill at baking. Enjoyment of buying food Spending time planning meals. The same class of homemakers gave less emphasis to (. 05 level of significance): Role Self Enjoyment of raising children. Enjoyment of interest in husband's work. It is difficult to derive a picture of this group from the above replies. The pattern above infers that the low expenditure group is that of the older homemaker without children, but this is not sub- stantiated in the relationships to follow. The class, low per capita expenditure for dairy products, was significantly associated with the following: Low per capita expenditures for food Low per capita mg. of thiamin in food purchased’l‘ Low per capita expenditures for: vitamin supplements, fats and oils, fruit. . 01 level of significance. 62 There were no relationships between the value classes for this variable and the following variables at the . 32 or less level of significance: Age of homemaker Education of homemaker Number of meals served at home Presence of children age 0-9, 10-19 Number of different items purchased Percent of income spent for food Per capita expenditures for: coffee, T. V. dinners Conflict score Social score All shopping attitudes except those oriented toward: price, information-seeking. ' There is a lack of a significant relationship between the value classes for this variable with those for per capita income and percent of income spent for food. There is, however, association of low per capita expenditure for food with low per capita expenditure for dairy products. This suggests low price and income elasticities for dairy products and that dairy products are a stable component of the diet in terms of expenditure levels. Note that there is a lack of relationship to age of homemaker and the presence of children of both age categories. 6 3 PER CAPITA EXPENDITURE FOR FATS AND OILS This variable was examined only with respect to the other variables. The class, low per capita expenditure for fats and oils, was significantly related to the following: Low per capita expenditure for food* Low age of homemaker Presence of children age 0-9 Per capita expenditure for: sugar, dairy products, fruits, * vegetables, * meats, * bakery products Per capita mg. of thiamin in food purchased. * Neither value class was related to the following variables at the . 32 or less level of significance: Education of homemaker Number of meals served at home Percent of different items purchased only once Per capita expenditure for: coffee, vitamin supplements, T. V. dinners Conflict score Social score All shopping attitudes except those oriented toward: price, husband, wife Percent of income spent for food Number of different items included in 50 percent of the food bill. . 01 level of significance. 64 This variable shows a stability toward measures of economic restriction on food purchases that is similar to that of dairy products. The low value class of this variable is associated with young home- makers with young children. PER CAPITA EXPENDITURE FOR FRUITS In replying to the role-self questionnaire the homemakers from the class, low per capita expenditure for fruit, gave more emphasis to (. 05 level of significance): Role Self Disciplining the children Skill at canning Enjoyment of sewing Disciplining the children Spending time cleaning house Spending time with children The same class of homemakers gave less emphasis to (. 05 level of significance): Role Self Mending clothes Being calm Being intelligent Being self—controlled Handling family finances. Planning meals. The pattern of replies above agrees with the relationships below indicating that the class, low per capita expenditure for fruits, to be under noticeable income restriction and to be the younger 65 homemaker with children. Seventy-one of the 118 families in the class, high per capita expenditure for fruits, were one or two person families. Only 23 families in the high value class had four or more persons. When examined with respect to the other variables the class, low per capita expenditure for fruits, was significantly associated with: Low per capita income’l< Low per capita expenditure for food* Low age of homemaker Presence of children, 0-9, 10-19 Low per capita expenditure for: sugar, * dairy products, fats and oils, vegetables, * meats, * bakery products* Low per capita dozen of eggs purchased Low social score Low shopping attitudes toward: quality, wife (dominant in shopping decisions) High percent of income spent for food. Neither value class for this variable was associated with the following variables at the . 32 or less level of significance: Education of homemaker Number of meals served at home Percent of different items purchased only once . 01 level of significance. 66 Per capita expenditures for: coffee, T. V. dinners Conflict score All shopping attitudes except those oriented toward quality, wife, husband Number of different items purchased Number of different items included in 50 percent of the food bill. The above relationship to both per capita income and the expenditure for food indicates that the demand for fruit should display rather high income elasticity. It should be kept in mind, however, that this expenditure data and the lower income families could be purchasing at quantity discounts. The increased emphasis on canning in the self concept supports this. This is the only food class with a significant association with quality oriented shopping attitudes. PER CAPITA EXPENDITURE FOR VEGETABLES This variable was examined only with respect to the other variables. The low value class was associated with the other variables as follows: Low per capita income Low per capita expenditure for food* Low age of homemaker Presence of children age 0—9* Low number of different items purchased . 01 level of significance. 67 Low number of different items included in 50 percent of the food bill Low per capita expenditures for: sugar, fats and oils, * fruits, * meats, * bakery goods* Low per capita mg. of thiamin in food purchased Low value, wife oriented shopping attitudes. Neither value class was associated with the following variables at the . 32 or less level of significance: Education of homemaker Presence of children age 10-19 Percent of different items purchased only once Expenditures for: coffee, T.V. dinners, vitamin supplements Conflict score All shopping attitudes except those oriented toward: quality, wife, husband. Comparison of the relationships of this variable to those of per capita expenditure for fruit suggests that there is less relationship between this variable and income and that the effect of the older children age 10-19 is indicated as they make adult-type demands under the lesser economic restrictions of the older wage earner. PER CAPITA EXPENDITURE FOR MEATS This variable was examined only with respect to the other variables. The low value class was significantly associated with: . 01 level of significance. 68 Low per capita income* Low per capita expenditure for food* Presence of children* Low per capita expenditure for: sugar, * fats and oils, * fruit, * vegetables, * bakery goods)? Low per capita dozen of eggs purchased’l‘ Low mg. of thiamin in food purchased. * Neither value class was associated with the following variables at the . 32 or less level of significance: Age, education of homemaker Number of meals served at home Presence of children age 10-19 Percent of income spent for food Percent of items purchased only once Number of different items included in 50 percent of the food bill Per capita expenditures for: vitamin supplements, T. V. dinners Conflict score Social score All shopping attitudes except those oriented toward price. The per capita expenditure for meats was highly related to per capita income and per capita expenditure for food but not related to percent of income spent for food at the . 32 or less level of significance or price oriented shopping attitudes at the . 10 or less level of significance. . 01 level of significance. 69 PER CAPITA EXPENDITURE FOR BAKERY GOODS This variable was examined only with respect to the other variables. The class, low per capita expenditure for bakery goods was significantly related to the following: Low per capita expenditure for food* Low number of meals served at home Presence of children age 0-9 Low per capita expenditures for: sugar, fruits, * vegetables, * meats, fats and oils. Neither value class for this variable was associated with the following variables at the . 32 or less level of significance: Per capita income Age of homemaker Percent of different items purchased only once Number of different items purchased Per capita expenditures for: coffee, vitamin supplements, T. V. dinners Per capita dozens of eggs purchased Conflict score Social score All shopping attitudes except those oriented toward price. The per capita expenditure for bakery products was associated with per capita expenditure for food but not with per capita income. . 01 level of significance. u. -< ‘v ayp, .nv - ’1) , o... _ i... M. “o. 7O Sub-sorting the population into groups homogeneous with respect to per capita income, and per capita expenditure for food, the above relationship was found to be the reflection of the presence of a class of 47 families with high per capita expenditure for food, high per capita expenditure for bakery goods, low per capita income. The group, high per capita income, low per capita expenditure for food, high per capita expenditure for bakery goods, was nonexistent. The group, low‘ per capita income, low per capita food expenditures, high per capita expenditures for bakery goods, contained 24 families. The largest group, 61 families, was characterized by joint membership in the low value classes for all three variables. The frequencies in these subgroups support the assertion that bakery goods figure sub- stantially in the higher levels of expenditure for food by low income families. This is also an example of the use of subgroup analysis. PER CAPITA MILLIGRAMS OF THIAMIN IN FOOD PURCHASED In replying to the role-self questionnaire the homemakers from the low value class gave more emphasis to (. 05 level of significance): Role Self (No significant differences. ) Disciplining the children Enjoyment of raising children Spending time with the children. 71 The same low value class gave less emphasis to (. 05 level of significance): Role Self Mending clothes. Running a quiet home Interest in husband‘s business. The above pattern is typically that of the young homemaker with young children in the household. The class, low per capita mg. of thiamin in food purchased, was significantly associated with the following: Low per capita income Low per capita expenditure for food* Low age of homemaker* Low number of meals served at home Presence of children age 0-9* Low per capita expenditure for: sugar, * vitamin supplements, dairy products, * fats and oils, * vegetables, ’3‘ meats, * bakery goods* Low per capita dozens of eggs purchased* Low score with respect to attitudes implying wife dominance in shopping decisions High education of homemaker. Neither value class was associated with the following variables at the . 32 or less level of significance: :3: . 01 level of significance. 012E . 5232 5 1‘. \'Fn Vou‘g s l *P .., ‘| be“ My.“ .,: Y ‘r ‘5“ 72 Percent of different items purchased only once Number of different items included in 50 percent of the food bill Per capita expenditure for: T. V. dinners, fruit All shopping attitudes except those oriented toward wife, husband. These relationships indicate that the amount of thiamin purchased is highly related to the expenditure levels for a broad group ”of foods and food classes, per capita income, and the per capita expenditure for food. The low score with respect to wife oriented shopping attitudes is typical of the young housewife with young children. Since thiamin requirements vary as a function of body weight, activity, and calory intake, no inference can be made from this data with respect to deficiency. CONFLICT SCORE In replying to the role-self questionnaire, homemakers with low conflict scores gave more emphasis to (. 05 level of significance): Role Self Housekeeping Serve a variety at meals Skill at canning Enjoy raising children Spending time cleaning house. Spending time with the children. The homemakers with low conflict scores gave less emphasis to (. 05 level of significance): U) T‘- Ix) ,..~ ‘V- 73 Role Self Helping in husband's job Running a cozy home Being ambitious Spending time cleaning house. Being charming. The above pattern resembles that of the housewife with young children in the home as is corroborated below. The homemakers with low conflict scores were from families associated with: Presence of children age 0-9 Low social score High education of homemaker. Neither value class was associated with the following variables at the . 32 or less level of significance. Per capita income Per capita expenditure for food Number of meals served at home Presence of children age 10-19 Percent of different items purchased only once Number of different items purchased Per capita expenditure for all foods wt: sugar, coffee Percent of income spent for food All shopping attitude variables Per capita mg. of thiamin in food purchased. II' L.) ' (I) SCI r.. u 'v.. ’ v 4., “ii. C1! ll: (‘1 1 74 This variable is characterized by a lack of relationship with respect to the sample as a whole. Since there is a relationship between conflict score and the presence of young children in the household, when conflict score is used to test hypotheses, it will be tested within groups homogeneous with respect to the presence or absence of children, age 0—9. SOCIAL SCORE The derivation of this variable has been discussed. It is, in general, a measure of the extent to which the homemaker's interests extend outside the home or to people outside of the immediate family. In replying to the role-self questionnaire, homemakers with a low role score gave more emphasis to (. 05 level of significance): Role Self Disciplining the children Running a comfortable home Housekeeping Disciplining the children Being a companion to the Being calm children Serving a variety at meals Enjoyment of raising children Skill at baking Spending time cleaning house. Enjoyment of buying food Enjoyment of raising children Spending time with the children. 75 The homemakers with low role scores gave less emphasis to (. 05 level of significance): Role Self Helping husband's career Running a quiet home Helping on husband's job Being informed on new ideas Keeping social responsibilities Socializing with the neighbors Skill at canning Interest in husband's work Enjoyment of budgeting finances Earning additional family income. Helping husband on job. This pattern indicates that those homemakers with low role scores stay at home, care for the children, and keep house. The homemakers with high role scores are occupied away from home, primarily with income earning activities, and .with social responsibilities. The class, low role score, was significantly associated with the following: Presence of children age 0-9 Low percent of different items purchased only once Low per capita expenditure for fruit Low conflict score. Neither value class was associated with the following variables at the . 32 or less level of significance: Per capita expenditure for food 76 Education of homemakers Number of meals served at home Presence of children age 10-19 Per capita expenditure for: coffee, sugar, T.V. dinners, dairy products Per capita dozens of eggs purchased. From these relationships one can only infer that the presence of younger children in the household hinders social and income earning activities by the homemaker. There is no significant relationship to the number of meals served at home. Using the implied relationship to high role score, it was observed that high per capita expenditure for fruit and high percent of total items purchased only once are associated with homemakers who have high role scores. PRICE ORIENTED SHOPPING ATTITUDES In replying to the role-self questionnaire the homemakers with low scores with respect to price oriented shopping attitudes gave more emphasis to being friendly in the role concept and less emphasis to skill at canning in the self concept (. 05 level of significance). There was a significant association between the low value class for price oriented shopping attitudes and the low value class for convenience oriented shopping attitudes. Neither value class for price oriented shopping attitudes was associated with the following variables at the . 32 or less level of significance: it“! ' m1 ..‘ \n i 77 Per capita income Age of homemaker Number of meals served at home Presence of children age 0-9, 10-19 Per capita mg. of thiamin in food purchased Per capita expenditures for: sugar, fruit Per capita dozens of eggs purchased Conflict score Social score Number of different items purchased Percent of income spent for food Number of different items included in 50 percent of the food bill All shopping attitudes except those oriented toward preplanning. This variable and the variable, convenience oriented shopping attitudes, are included in the subgroup analysis. Upon viewing the questions from which the values for these variables are derived, it will be seen that the general tenor of the questions implies a fair degree of decision-making. Differences between high and low value classes for this variable are assumed to indicate differences in willingness to make decisions. CONVENIENCE ORIENTED SHOPPING ATTITUDES In replying to the role-self questionnaire, homemakers in the low value class gave more emphasis to (. 05 level of significance): 78 Role Self Skill at baking Running a quiet home Skill at canning. Enjoyment of shopping for food. The same class of homemakers gave less emphasis to (. 05 level of significance): Role Self Budgeting family finances. Socializing with the neighbors. There was a significant association between the low value class for this variable and: Low value, shopping attitudes oriented toward: preplanning, information-seeking, calculation, other (people), wife. Neither value class for this variable was associated with the following variables at the . 32 or less level of significance: Per capita income Per capita expenditure for food Age, education of homemaker Number of meals served at home Presence of children, age 0-9, 10-19 Percent of different items purchased only once Number of different items purchased Percent of income spent for food Number of different items included in 50 percent of the food bill Per capita expenditures for all food classes except: coffee, vitamin supplements 79 Per capita dozens of eggs purchased Conflict score Role score. Using the implied statement about the high value class for this variable, it was found that homemakers who were highly oriented toward convenience in their shopping attitudes were also highly oriented toward preplanning, information-seeking and calculation. There is no significant relationship to the socio-economic variables. The remaining shopping attitude variables will not be listed separately and were not examined with respect to the role-self question- naire. It was found in general that preplanning, information-seeking, and calculation oriented shopping attitudes were interrelated with each other as were the questions from which they were derived. 80 CHAPTER III SUB GROUP ANALYSIS This chapter focuses on certain well-defined subgroups within the sample of 211 families investigated in the previous chapter. Significant relationships were sought between specified variables within subgroups that were homogeneous with respect to several socio- economic variables. The preceeding results of examining all possible pairs of variables for significant relationships, indicated that per capita income, age of homemaker, per capita expenditure for food, percent of income spent for food, and the presence of children age 0-9, are highly inter- related. In addition it was noted that the conflict score was related to the presence of children 0—9. In the subgroup analysis to follow, conflict score and per capita milligrams of thiamin in the food purchased are examined rather closely. To isolate the relationship between these and other variables from the simultaneous intercorrelation between socio—economic variables, subgroups were formed which were homo- geneous with respect to these interrelated socio-economic variables. The procedure used was to sort the families with respect to the socio-economic variables to be held constant and discard those sub- groups containing fewer than 10 families. The following pairs of relationships were then examined: 81 Conflict score in relation to: Percent of different items purchased only once Number of items included in 50 percent of the food bill Convenience oriented shopping attitudes Price oriented shopping attitudes Per capita dozens ofeggs purchased Per capita expenditures for: sugar, fruit, dairy products. Per capita mg. of thiamin in food purchased in relation to: Percent of items purchased only once Number of different items included in 50 percent of the food bill Conflict score Convenience oriented shopping attitudes Price oriented shopping attitudes. Two kinds of tests for relationships in the subgroup analysis are reported. The first is for the pairs of relationships above. The second is a report for differences in the values of the variables listed above between two subgroups that differ by only one socio-economic variable. To phrase this differently, conflict score and the level of thiamin purchased were examined within the subgroups. These variables and the others listed above were observed in relation to socio-economic differences between subgroups. The following specific hypotheses derived from the initial questions in this thesis were tested: 82 1. There are significant relationships between conflict score and (a) shopping behavior variables, (b) attitudes toward shopping, and (c) per capita expenditures for specific food classes. 2. There are significant relationships between the level of thiamin intake and (a) conflict score, (b) shopping attitudes, and(c) shopping behavior variables. The relationships sought in observing differences between groups that differed only with respect to one socio-economic variable, are important to demand analysis and in relating the thesis to economic theory. Specifically the differences in per capita expenditure levels will be noted between groups that differed with respect to one of the socio-economic variables for which each subgroup is homogeneous." In this manner the expenditure levels for dairy products, fruit, sugar, and the quantity of eggs purchased may be associated with the value classes for per capita income, age of homemaker, per capita expenditure for food, or the presence of children age 0-9. Significance was established by means of the chi-square test, with 1 degree of freedom. RELATIONSHIPS WITHIN SUBGROUPS In the group, low per capita income, high age of homemaker, high per capita expenditure for food, no children age 0-9 (30 families): Low conflict was associated with high convenience oriented shopping attitudes (. 10 level of significance). Low per capita mg. of thiamin in food purchased was associated with high convenience oriented shopping attitudes (. 05 level of significance). 83 In the group, high per capita income, low age of homemaker, high per capita expenditure for food, no children age 0-9 (16 families): Low conflict score was associated with low number of different items included in 50 percent of the food bill (. 10 level of significance). In the group, low per capita income, high age of homemaker, high per capita expenditure for food, no children age 0-9, with the added restriction to families of size 1. 5-2. 5 average persons (18 families): Low conflict score was associated with high convenience oriented shopping attitudes (. 10 level of significance). Low per capita mg. of thiamin in food purchased was associated with low conflict score (. 10 level of significance). In the group, high age of homemaker, high per capita expenditure for food, low percent of income spent for food, no children age 0-9 (37 families): Low per capita mg. of thiamin in the food purchased was associated with low price oriented shopping attitudes (. 05 level of significance). RESULTS OF ANALYSIS OF SUBGROUPS WITH RESPECT TO SOCIO-ECONOMIC VARIABLES Within the group low per capita income, low age of homemaker, children present age 0—9 (58 families): Low per capita expenditure for food was associated with: Low per capita dozens of eggs purchased’l‘ Low per capita expenditure for dairy products* . 01 level of significance. 84 Low per capita expenditure for sugar and confections** Low per capita expenditure for fruit (. 05 level of significance). Within group, low per capita income, high age of homemaker, no children age 0-9 (59 families): Low per capita expenditure for food was associated with: High percent of items purchased only once** Low per capita expenditure for fruit. ** Within group, high age of homemaker, high per capita expenditure for food, no children age 0-9 (67 families): Low per capita income was associated with: Low percent of different items purchased only once (. 10 level of significance). Within group, high per capita income, high per capita expenditure for food, no children 0-9 (53 families): Low age of homemaker was associated with: High convenience oriented shopping attitudes. * Within group, low per capita income, high age of home- maker, no children age 0-9, size of family, 1. 5-2. 5 average persons (38 families): Low per capita expenditure for food was associated with: ‘ High number of items purchased only once (. 05 level of significance) Low per capita expenditure for fruit. ** Within group, high age of homemaker, high per capita expenditure for food, no children age 0-9, size of family 1. 5-2. 5 . 01 level of significance. . 001 level of significance. . I ‘l-,- 4 r- . , ”I“ “w“ .- ~ . :vo ., "-... 0‘ v ‘- ‘4' ._A v .v-o -.‘ .., v .- , _.-. '— ‘- .\~ ‘s (I, 85 average persons (39 families): Low per capita income was associated with: High conflict score (. 02 level of significance). Between groups homogeneous with respect to high per capita income, high age of homemaker, high per capita expenditure for food, no children age 0-9, but differing with respect to family size, less than 1. 5 average persons vs. 1. 5-2. 5 average persons, the smaller size of family was associated with the high conflict score. * Within group, low age of homemaker, low percent of income spent for food, low per capita expenditure for food (42 families): The presence of children age 0-9 was associated with: Low per capita expenditure for fruits (. 10 level of significance) Low conflict score (. 10 level of significance). Within group, low age of homemaker, low percent of income spent for food, no children age 0-9 (35 families): Low per capita expenditure for food was associated with: Low per capita expenditure for sugar and confections. * Within group, low age of homemaker, high percent of income spent for food, children age 0-9 present (31 families): Low per capita expenditure for food was associated with: Low per capita expenditure for dairy products (. 02 level of significance) Low per capita expenditure for sugar and confections. ** . 01 level of significance. . 001 level of significance. ...,.,. '5.... nu... "’.s- N .4 O‘ h. s.‘ ‘.. 86 Within group, high age of homemaker, low percent of income spent for food, no children age 0-9 (54 families): Low per capita expenditure for food was associated with: Low per capita expenditures for dairy products** Low per capita expenditure for sugar (. 10 level of significance) Low per capita expenditure for fruits. ** Within group, high age of homemaker, high percent of income spent for food, no children age 0-9 (44 families): Low per capita expenditure for food was associated with: High percent of items purchased only once* Low per capita dozens of eggs purchased* Low per capita expenditures for dairy products (. 05 level of significance) Low per capita expenditure for fruits. ** Within group, high age of homemaker, high per capita expenditure for food, no children age 0—9 (67 families): Low percent of income spent for food was associated with: High percent of different items purchased only once. * These relationships in the sections dealing with both the analysis within groups with respect to conflict score, and the level of thiamin intake, and between socio-economic subgroups is a list of all relationships existent for the given set of variables and groups of size more than 10 families at the . 10 or less level of significance. . 01 level of significance. . 001 level of significance. 87 SUMMARY OF FINDINGS 1. Per capita expenditure for food was widely and significantly related to the food expenditure variables as would be anticipated. 2. High percent of items purchased only once was associated with: Low per capita expenditure for food (in 3 groups) High per capita income Low percent of income spent for food. The relationship of high percent of items purchased only once to low per capita expenditure for food is in contradiction to the observations in the analysis of the whole sample. It will be noticed, however, that this was observed in groups in which the age of home- maker was high and the per capita income low, or percent of income ': spent for food was high. It may well be that for this subgroup of older homemakers, a high percent of items, purchased only once, is a reflection of price sensitivity and ”close shopping. " 3. Low conflict score was associated with: Low number of different items included in 50 percent of the food bill High convenience orientation Low mg. of thiamin in food purchased High per capita income Two person rather than single person families Presence of children age 0-9. Hypotheses (p. 82), . l (a), l (b), and 2 (a) are supported. 88 There is no support for the hypothesis that conflict score is related to level of expenditure for specific foods or food classes. 4. Low per capita mg. of thiamin in food purchased was related to: High convenience orientation Low conflict score Low price orientation. Hypotheses (p. 82 ) 2 (a) and 2 (b) are supported. No relationship was found between level of thiamin in food purchased and shopping behavior variables. 5. High convenience orientation in shopping attitudes was related to: Low conflict score Low mg. of thiamin in food purchased Low age of homemaker. 6. High price orientation was related to: High per capita mg. of thiamin in food purchased. The above complex of relationships runs contrary to a line of reasoning that asserts that high conflict score is indicative of a condition such that there is reluctance to make decisions or that it is associated with the lower thiamin levels. An inference that does bring the whole complex of relationships into a consistent framework is the assertion that high role conflict is an active decision state or that low role conflict is the result of a state such that a person is highly 89 convenience oriented, shows a low orientation toward price calculation, includes only a low number of items in 50 percent of the food bill, is on a lower level of thiamin intake, under income restriction, and pushed for time with young children in the home. This scheme is highly conjectural but at least consistent. ,."' .\ .a~'" - .H' ,. p b .o“. .- -.. ‘ .v' p .4 n"‘ )A nufi‘“ , . ...-— " II“.‘ .., 1‘ - Inch \— . .‘r‘ o a 90 CHAPTER IV SUMMARY AND CONCLUSIONS The problem area was initially stated in the form of the question: Does a relationship exist among, (a) observed purchasing behavior, (b) verbalized shopping attitudes, (c) role and self perceptions of the homemaker, (d) socio-economic characteristics of the family, and (e) expenditures for quantities of foods or food classes purchased? Examples of relationships of all possible pairs of the above were found at levels of . 10 or less significance. Itwas possible to relate specific answers of the role-self questionnaire to specific characteristics of families. These patterns of significant replies form profiles of the families in each food expenditure class, and those families exhibiting particular patterns of food purchasing behavior and attitudes toward purchasing food. In the subgroup analysis two hypotheses were formed from the assertions made in Chapter I. 1. Conflict score is significantly related to: (a) shopping behavior, (b) attitudes toward shopping, and (c) per capita expenditures for specific food classes. 2. Per capita of thiamin in the food purchased is significantly related to: (a) conflict score, (b) shopping attitude, (c) shopping behavior. 91 Hypotheses l, (a), (b) and 2, (a), (b), were supported. Hypotheses l (c), and 2 (c) were not supported at the .10 level of significance. The relationship between the analyses conducted in this thesis and the general investigation of the nature of the consumers of food is plain. The decision-energy concept is more subtle but no less important. The equimarginal or indifference approaches to the explanation of consumer behavior assert that the consumer will respond to small changes in information received concerning proposed allocation of personal resources. If there is a limit to the capacity to evaluate incoming information, or a refusal to deal with incoming information, conscious or unconscious, the failure to respond results in what is normally termed irrational behavior. If the feelings of well-being affect the willingness to assess incoming information concerning food, and this feeling of well-being is a function of the nutritional content of the food, a circular relationship results. This may or may not constitute an area for fruitful investigation, by home economists. Apart from this, however, the investigation of habit and partial use of information in decision-making should be fruitful in forming more realistic decision models. An important point to be made in the light of the preceeding study is that the life cycle generates distinct types of families. The sorting procedures used in this study, when carried to formation of subgroups homogeneous with respect to 4 variables, typically yielded 92 5 or 6 groups of size, 10 families or more, and the remaining 9 or 10 possible subgroups either were nonexistent or had few families in them. Sorting to 2 or 3 additional variables generally eliminated all but 2 groups of sufficient size within the sample of 211 families to yield meaningful data. It can be observed that analyses that combine groups of families from divergent periods in the life cycle can be misleading. The case of the cancellation of attitudes between younger and older homemakers within a given income class was discussed. The recommendation is that panels designed to evaluate consumer attitudes or psychological relationships should either be very large or confined to a few well-defined groups of families; the families in each group being very similar with respect to a large number of socio- economic characteristics. There is a comment to be made concerning the binary aggregation system used in this study. The statement can be made that such a bivalued system describes a family in terms that say that it is, or is not, a member of a class of families of given characteristics. Each family is an individual unit and is recognized as such and the aggregation process deletes those characteristics which are estimated to have no bearing on the relationship under investigation. The bi- valued system used in this study is an effective technique for checking which characteristics are in fact not related to the investigation. It is an effective device for handling nonquantifiable variables since 93 assignment of value requires only consistent identification. The technique is limited in usefulness as a precise prediction device. Looking forward, the next step in investigation of the decision energy concept is the drawing together of information that has been developed within several disciplines. The idea of decision energy lies close to the attention-getting and holding that preoccupy both the education and advertising world. There is work in existence relating to habits. There are some data existent in panels that probe the interpretation of consumer grades by consumers, that, although designed for other purposes, yield good information, obtained under closely controlled conditions. There is a body of information theory that can be brought to bear on the problem. It is obvious that the line between such investigation and the disciplines of psychology, communications, and political science are not well-defined, indicating that the implications and applications of such study are broad. Within economics the applications are not confined to management of personal resources but are of relevance to the management aspects of production and marketing, and the communications aspects of policy and extension activities. APPENDIX 94 95 DECISION-MAKING SCALE BY CATEGORY Price oriented: One who bases his food purchase decisions primarily on the cost of the item, irrespective of the time needed for preparation, the taste of the item, or a pre-established habit of food eating. 30. I read the newspapers to find which food stores are having specials and I shop at those stores which are having an attractive offer. 41% (1) a. Rarely or almost never true 12% (2) b. Sometimes but infrequently true 20% (3) c. Occasionally true 12% (4) d. Very often true 15% (5) e. True most of the time 31. I follow the prices of foods very closely. I know when the basic foods have either increased or decreased even by only a few cents. 16% (l) a. Rarely or almost never true 9% (2) b. Sometimes but infrequently true 11% (3) c. Occasionally true 24% (4) (1. Very often true 40% (5) e. True most of the time 39. Often before I buy a certain brand of food item, I compare the various sizes to determine the actual cost per ounce of each size. 28% (l) a. Rarely or almost never true 10% (2) b. Sometimes but infrequently true 19% (3) c. Occasionally true 22% (4) d. Very often true 21% (5) e. True most of the time 27. Before I go to the store I figure which foods I can buy that will cost the least amount of money. 22% (1) a. Rarely or almost never true 15% (2) b. Sometimes but infrequently true 21% (3) c. Occasionally true 20% (4) d. Very often true 22% (5) e. True most of the time 9. I buy something else when the price of a food item I usually buy goes up. 96 18% (1) a. Rarely or almost never true 15% (2) b. Sometimes but infrequently true 22% (3) c. Occasionally true 23% (4) d. Very often true 22% (5) e. True most of the time Quality oriented: One who bases his food purchase decisions primarily on the appearance, flavorfulness, healthfulness, and kinesthetic qualities of the items considered regardless of cost, or facility of preparation. 1. 10. In buying foods I figure the amount of calories and nutrients they contain. 44% (l) a. Rarely or almost never true 12% (2) b. Sometimes but infrequently true 16% (3) c. Occasionally true 14% (4) d. Very often true 14% (5) 6. True most of the time The food decisions I make before going to the store are based primarily on the flavorfulness and healthfulness of the food. 13% (l) a. Rarely or almost never 9% (2) b. Sometimes but infrequently true 16% (3) c. Occasionally true 26% (4) d. Very often true 36% (5) e. True most of the time I listen to the information services offered on the radio to find out which foods are good and nutritious. 46% (l) a. Rarely or almost never 14% (2) b. Sometimes but infrequently true 19% (3) c. Occasionally true 12% (4) d. Very often true 9% (5) e. True most of the time I can tell very quickly when the flavor, freshness, or appearance of food that I have been buying regularly changes. 5% (l) a. Rarely or almost never 6% (2) b. Sometimes but infrequently true 12% (3) c. Occasionally true 31% (4) d. Very often true 46% (5) e. True most of the time z 0 II O I a- JC ‘2- J1 (A) l" (.7. t‘v D—- k») H (II '_..« ._. . i. n”! PM: \- *A 97 32. When I buy food, I buy the very best quality no matter what the price. 22% 22% 23% 18% 15% (l) a. (2) b. (3) c. (4) d. (5) e. Rarely or almost never true Sometimes but infrequently true Occasionally true Very often true True most of the time Convenience oriented: One who bases his food purchase decisions primarily on the time and facility of preparation regardless of the cost or quality of the item. 8. I look in the food pages of the newspapers for food items that can be quickly and easily prepared. 37% (l) a. Rarely or almost never true 15% (2) b. Sometimes but infrequently true 22% (3) c. Occasionally true 14% (4) d. Very often true 12% (5) e. True most of the time 24. I compare instructions on various food packages so that I can calculate which food will be easiest and quickest to prepare. 35% (l) a. Rarely or almost never true 16% (2) b. Sometimes but infrequently true 23% (3) c. Occasionally true 15% (4) (:1. Very often true 11% (5) e. True most of the time 28. I figure out in advance before going to the store foods and meals that can be quickly and easily prepared. 28% (1) a. Rarely or almost never true 13% (2) b. Sometimes but infrequently true 35% (3) c. Occasionally true 16% (4) d. Very often true 8% (5) e. True most of the time 35. I like to spend as little time as possible preparing meals, that is why I buy the good, frozen, ready-made dishes that can be prepared quickly. 59% (l) a. Rarely or almost never true 17% (2) b. Sometimes but infrequently true 16% (3) c. Occasionally true 6% (4) (1. Very often true 2% (5) e. True most of the time " . d‘;31¢4 Dov A“ . ‘ I i I e "u. 4C it nL \IL «)9 a). fili- 98 36. I seldom spend more than 30 minutes in preparing the day‘s largest meal. 51% (l) a. Rarely or almost never true 21% (2) b. Sometimes but infrequently true 12% (3) c. Occasionally true 9% (4) d. Very often true 7% (5) e. True most of the time Preplanning oriented: One who, prior to reaching the site of purchase, knows what foods he is going to purchase and how and when such foods will be used. 3. The food decisions I make before going to the store are based primarily on the flavorfulness and healthfulness of the food. 13% (l) a. Rarely or almost never true 9% (2) b. Sometimes but infrequently true 16% (3) c. Occasionally true 26% (4) (1. Very often true 36% (5) e. True most of the time 6. Before I go to the market I make out a complete grocery list. 22% (l) a. Rarely or almost never 6% (2) b. Sometimes but infrequently true 12% (3) c. Occasionally true 13% (4) d. Very often true 47% (5) e. True most of the time 23. I plan my menu, meal for meal, a couple of days in advance. 32% (l) a. Rarely or almost never true 16% (2) b. Sometimes but infrequently true 16% (3) c. Occasionally true 12% (4) d. Very often true 24% (5) e. True most of the time 2.7. Before I go to the store I figure which foods I can buy that will cost the least amount of money. 22% (1) a. Rarely or almost never true 15% (2) b. Sometimes but infrequently true 21% (3) c. Occasionally true 20% (4) d. Very often true 22% (5) e. True most of the time 28. 99 I figure out in advance before going to the store foods and meals that can be quickly and easily prepared. 28% 13% 35% 16% 8% (l) a. (2) b. (3) c. (4) d. (5) e. Rarely or almost never true Sometimes but infrequently true Occasionally true Very often true True most of the time Information-seeking oriented: One who actively and purposefully seeks facts from various agencies and services pertaining to price, quality, or convenience of various food items. 7. 13. 29. I listen to the information services offered on the radio to find out which foods are good and nutritious. 46% 14% 19% 12% 9% (l) a. (2) b. (3) c. (4) d. (5) e. Rarely or almost never true Sometimes but infrequently true Occasionally true Very often true True most of the time I look in the food pages of the newspapers for food items that can be quickly and easily prepared. 37% 15% 22% 14% 12% (l) a. (2) b. (3) c. (4) d. (5) e. Rarely or almost never true Sometimes but infrequently true Occasionally true Very often true True most of the time I read many magazine articles concerning foods and meals. 18% 11% 19% 28% 24% (1) a. (2) b. (3) c. (4) d. (5) e. Rarely or almost never true Sometimes but infrequently true Occasionally true Very often true True most of the time I read Consumer's Union, Consumer‘s Research, Changing Times, government publications or consumer's service publications to get ideas on buying foods. 71% 11% 10% 4% 4% (l) a. (2) b. (3) c. (4) d. (5) e. Rarely or almost never true Sometimes but infrequently true Occasionally true Very often true True most of the time 30. 100 I read the newspapers to find which food stores are having specials and I shop at those stores which are having an attractive offer. 41% (1) a. Rarely or almost never true 12% (2) b. Sometimes but infrequently true 20% (3) c. Occasionally true 12% (4) d. Very often true 15% (5) e. True most of the time Calculation oriented: One who carefully weighs and measures all the information that he has in order to aid him in making a food purchase decision. 1. 17. 24. 26. In buying foods I figure the amount of calories and nutrients they contain. 44% (l) a. Rarely or almost never true 12% (2) b. Sometimes but infrequently true 16% (3) c. Occasionally true 14% (4) d. Very often true 14% (5) e. True most of the time When I go food shopping I take time and use a pencil and paper or some other device to aid in figuring. 69% (l) a. Rarely or almost never true 8% (2) b. Sometimes but infrequently true 9% (3) c. Occasionally true 4% (d) d. Very often true 10% (5) e. True most of the time I compare instructions on various food packages so that I can calculate which food will be easiest and quickest to prepare. 35% (1) a. Rarely or almost never true 16% (2) b. Sometimes but infrequently true 23% (3) c. Occasionally true 15% (4) d. Very often true 11% (5) e. True most of the time The meat that I decide to buy is often determined by the number of servings and meals which I have figured it will supply. 12% (l) a. Rarely or almost never true 5% (2) b. Sometimes but infrequently true 10% (3) c. Occasionally true 33% (4) d. Very often true 40% (5) e. True most of the time 101 39. Often before I buy a certain brand of food item, I compare the various sizes to determine the actual cost per ounce for each size. 28% 10% 19% 22% 21% (l) a. (2) b. (3) c. (4) d. (5) e. Rarely or almost never true Sometimes but infrequently true Occasionally true Very often true True most of the time Other oriented: One whose food purchase decisions, practices, and philosophies are influenced by neighbors, friends or associates. 2. I have received some excellent ideas about food buying from our friends. 42% (l) a. Rarely or almost never true 20% (2) b. Sometimes but infrequently true 19% (3) c. Occasionally true 15% (4) d. Very often true 4% (5) e. True most of the time 15. The meals that we eat are very similar to the meals that our friends eat. 30% 15% 25% 21% 9% (l) a. (2) b. (3) c. (4) d. (5) e. Rarely or almost never true Sometimes but infrequently true Occasionally true Very often true True most of the time 18. I pretty well know what my friends like or dislike. 20% (l) a. Rarely or almost never true 14% (2) b. Sometimes but infrequently true 15% (3) c. Occasionally true 26% (4) d. Very often true 25% (5) e. True most of the time 14. Conversations with my friends have changed some of my food buying habits. 45% 18% 25% 9% 3% (1) a. (2) b. (3) c. (4) d. (5) e. Rarely or almost never true Sometimes but infrequently true Occasionally true Very often true True most of the time 34. 102 My friends and I discuss menus, meals, and foods with each other. 18% (l) a. Rarely or almost never true 18% (2) b. Sometimes but infrequently true 32% (3) c. Occasionally true 23% (4) (1. Very often true 9% (5) e. True most of the time Parent oriented: One whose food purchase decisions, practices, and philosophies are influenced by his or his spouse's parents. 5. 12. 19. 25. My parents have given my many recipes and/ or other ideas on food buying. 43% (1) a. Rarely or almost never true 14% (2) b. Sometimes but infrequently true 16% (3) c. Occasionally true 15% (4) (1. Very often true 12% (5) e. True most of the time My spouse's parents have given me many recipes and/ or other ideas on food buying. 63% (l) a. Rarely or almost never true 9% (2) b. Sometimes but infrequently true 13% (3) c. Occasionally true 10% (4) d. Very often true 5% (5) e. True most of the time The meals that my family eats are very similar to the meals that my spouse's parents eat or ate. 45% (1) a. Rarely or almost never true 17% (2) b. Sometimes but infrequently true 16% (3) c. Occasionally true 12% (4) d. Very often true 10% (5) e. True most of the time The meals that my family eats are very similar to the meals that my parents eat or ate. 27% (l) a. Rarely or almost never true 15% (2) b. Sometimes but infrequently true 25% (3) c. Occasionally true 19% (4) d. Very often true 14% (5) e. True most of the time 33. The size of the meals and the kinds of food that we eat are entirely different from those of ‘our parents. 30% (1) a. Rarely or almost never true 14% (2) .b. Sometimes but infrequently true 28% (3) c. Occasionally true 13% (4) d. Very often true 15% (5) e. True most of the time Dominance: One who makes the decisions on where to shop, what to buy, and how much money to spend for food. Wife 22. In our home, I am the boss of the kitchen and how much and what foods are purchased is my concern. 14% (1) a. Rarely or almost never true 4% (2) b. Sometimes but infrequently true 10% (3) c. Occasionally true 16% (4) (1. Very often true 56% (5) e. True most of the time 37. I decide how much money our family can or will spend for food. 30% (1) a. Rarely or almost never true 5% (2) b. Sometimes but infrequently true 9% (3) c. Occasionally true 15% (4) d. Very often true 41% (5) e. True most of the time 103 38. My husband wants nothing to do with the food buying or deciding what to eat. 40% (l) a. Rarely or almost never true 12% (2) b. Sometimes but infrequently true 14% (3) c. Occasionally true 10% (4) (1. Very often true 24% (5) e. True most of the time Husband 4. My husband is the business man in this family. I let him work out the budget and amount we have available for food. 16. 20. 104 61% (1) a. Rarely or almost never true 6% (2) b. Sometimes but infrequently true 7% (3) c. Occasionally true 8% (4) d. Very often true 18% (5) e. True most of the time My husband tells me how much money can be spent for food. 71% (l) a. Rarely or almost never true 7% (2) b. Sometimes but infrequently true 7% (3) c. Occasionally true 4% (4) d. Very often true 11% (5) e. True most of the time My husband tells me what I should buy at the store. 78% (l) a. Rarely or almost never true 8% (2) b. Sometimes but infrequently true 10% (3) c. Occasionally true 3% (4) d. Very often true 1% (5) e. True most of the time 105 BIBLIOGRAPHY Bilkey, Warren J. The Basic Relationships in Consumer Expenditure Behavior. Harvard Studies in Marketing Farm Products No. 4-H, Cambridge, Massachusetts, October, 1951. Hershey, Gerald. ”College Grades in Relation to Inventory Measures of Personality. " Unpublished Master's thesis, Michigan State University, 1958. Katona, George. Psychological Analysis of Economic Behavior. New York: McGraw-Hill Book Co. , Inc. , lst Ed. , 1951. Mowrer, Harriet. Personality Adjustment and Domestic Discord. New York: American Book Company, 1935. Norris, Ruby Turner. The Theory of Consumer's Demand. New Haven: Yale University Press, Rev. Ed. , 1952. Orient-Keiles, E. and Hallmann, L. F. The Breakfast Meal in Relation to Blood Sugar Levels. U. S.D.A. Circular 827, 1949. Schreier, Fred T. Human Motivation. Glencoe, Illinois: The Free Press, 1957. Snygg, D. and Combs, A. W. Individual Behavior. New York: Harper 8: Bros. , 1949. Trier, Howard E. ”Sociological Variables, Personality Traits, and Buying Attitudes Related to Role Perceptions and Conflicts: Among 242 Michigan Housewives. " Unpublished Ph. D. thesis, Michigan State University, 1959. White, Robert W. The Abnormal Personality. New York: The Ronald Press Co. , 1956. Williams, R. D. , et a1. ”Induced Thiamin (Vitamin B ) DefiCienCY in Man, " Archives of Internal Medicine, LXIX, 1942. 105 Pane: No . PART I The following is a list of 39 statements about food buying. After each question are the numbers from 1 to 5. Please indicate the degree to which the statement applies to you and your family by circling the number according to the following: 10. 11. 12. 13. 14. 1 - means rarely or never true 2 - means sometime but not often true - means occasionally true means very often true - means true most of the time 3 4 5 In buying foods I figure the amount of calories and nutrients they contain. ‘ 1 2 3 4 S I have received some excellent ideas about food buying from our friends. 1 2 3 4 5 The food decisions I make before going to the store are based primarily on the flavorfulness and healthfulness of the food. 1 2 3 4 5 My husband is the business man in this family, I let him work out the budget and amount we have available for foods. 1 2 3 4 5 My parents have given me many receipes and/ or ideas on food buying. 1 2 3 4 5 Before I go to the market I make out a complete grocery list. 1 2 3 4 5 I listen to the information services offered on the radio to find out which foods are good and nutritious. l 2 3 4 5 I look in the food pages of the newspapers for food items that can be quickly and easily prepared. 1 2 3 4 5 I buy something else when the price of a food item I usually buy goes up. 1 2 3 4 5 I can tell very quickly when the flavor, freshness, or appearance of a good food that I have been buying regularly changes. 1 2 3 4 5 My spouse's parents have given me many receipes and/or ideas on food buying. 1 2 3 4 5 I read many magazine articles concerning foods and meals. 1 2 3 4 5 Conversations with my friends have changed some of my food buying habits. l 2 3 4 5 The meals that we eat are very similar to the meals that our friends eat. 1 2 3 4 5 I. i . A x... ‘ .. fl n q n r 1‘ py n . I :I‘ a. a . . ~ A I . y .. i. u v. . e . Q . o . I . mat . v I. x . . . v . 4 . . . . . . ~ . . n. A . . i. .. . v.1 . o . A. . . . . 1‘ . . . . a, . . .. .. . . . . . . . u . .- . . . i . . .. . . ‘ q . ‘ .. V I . r . . e . . . . . . . y u . . u. . A . . ,3 . o n c t . | ,. . t r. . . . a . . . . ~ . A . . . J». . . I . . .o . .. . v . o . o n . . y . .. i .. . u. . v u. . . u . .1 n . . . . .v u. . .. .. n . . . . o A o . u A . i u . . ll . . y . . v . . . y. ‘ . . ._ . . , A . . 9 t . . A . t .- . . . s ,, . v .. A . . . . . . v . . . .A V. \ . r 1 . .uo - . a . ‘ . L . . . . A . , . . \ . . . _ . t. . _ ' . a . . . 9 . I 1 . . . . ‘_ ~ V V . u . . .. c . A . .» . . . . . . . a . . I . . . 4 .1 .. . . A . . . . . . . . . O. ‘ . . . . . . 4 a . . . A . . I r . ‘ t o . ; A. . . ~ . . . . . c n . . . . a . . . . 4 . 9. a u . . 1 . . n . . . . . .p . v . . .o .A. . o . . . A . . . . f . A . . . . n .. A! s . A . c .. . . . a . . I .. . . o... . .. . u . . , I p ) , .. w P . . . g . .. . u a . . \ . . c o . t . ~ u r c. u. . . . . u .. , . . p _ . l a. I .- 0 . ~ . . . . . . . c . I 1! u . . .\ A o 1.. s r. . . . . , .. a. ‘ . . . ,C. . .. . . . u A n. . . 54 A. . .. .r I a . A» . . u . .. o . . t . . a .u u], . . A .. . . . A n . . . u . A . . . . pvt I . . ‘ . . . o . . a . . . . . a . . : 6. .1 . u . . .. . . . a r u . . . . ... I ..A . . . . . o vV. . A. I . . p. I . .. n . .. I. A . A u . . . .. u . . . ‘ v . . 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 106 -2- - means rarely or never true - means sometimes but not often true means occasionally true - means very often true means true most of the time Ut-L‘wNn-I I My husband tells me how much money can be spent for food. 1 2 3 4 5 when I go food shepping I take and use a pencil and paper or some other device to aid in figuring. 1 2 3 4 5 I pretty well know what foods my friends like or dislike. l 2 3 4 S The meals that my family eats are very similar to the meals that my spouse's parents eat or ate. . 1 2 3 4 5 My husband tells me what I should buy at the store. 1 2 3 4 5 In our home, I am the boss of the kitchen and how such and what foods are purchased is my concern. 1 2 3 4 5 I plan my menu, meal for meal, a couple of days in advance. 1 2 3 4 5 I compare instructions on various food packages so that I can calculate which food will be easiest and quickest to prepare. 1 2 3 4 5 The meals that my family eats are very similar to the meals that my parents eat or ate. 1 2 3 4 5 ' The meat that I decide to buy is often determined by the number of servings and meals which I have figured it will supply. 1 2 3 4 5 Before I go to the store I figure which foods I can buy that will cost the least amount of money. 1 2 3 4 5 I figure out in advance before going to the store foods and meals that can be quickly and easily prepared. 1 2 3 4 S I read Consumers Union, Consumers Research, Changing Times, government publications, or consumer service publications to get ideas on buying foods. ~ 1 2 3 4 5 I read the newspapers to find which food stores are having specials and I shOp at these stores which are having an attractive offer. 1 2 3 4 5 I follow the prices of foods very closely. I know when the basic foods have either increased or decreased even by only a few cents. I 2 3 4 5 When I buy food, I buy the very best quality no matter what the price. 1 2 3 4 5 The size of the meals and the kinds of foods that we eat are entirely differ- ent from those of our parents. 1 2 3 4 S ' _. L‘L—<—;‘ (#— 32. 33. 34. 35. 36. 37. _ 3 . 107 1 - means rarely or never true 2 - means sometimes but not often true 3 - means occasionally true 4 - means very often true 5 - means true most of the time My friends and I discuss menus, meals and foods with each other. 1 2 3 4 S I like to spend as little time as possible preparing meals, that is why I buy the good, frozen, ready-made dishes that can be prepared quickly. 1 2 3 4 5 I seldom spend more than 30 minutes in preparing the days largest meal. ' 1 2 3 4 5 I decide how such our family can or will spend for food. 1 2 3 4 5 My husband wants nothing to do with the food buying or deciding what to eat. 1 2 3 4 5 Often before I buy a certain brand of food item, I compare the various sizes to determine the actual cost per ounce for each size. 1 2 3 4 5 108 PART II Instructions: Following are lists of ”adjectives or phrases describing behaviors and characteristics of a homemaker. From each group of words or phrases we would like you to select the most important four which picture how you feel a handmaker under economic circumstances similar to your own should behave. I What kind of a home should the homemaker run? Pick ’4 __ cheerful cozy relaxed ....... efficient —' clean - fashionable \. '— modern - comfortable ‘1 : orderly : quiet II. what should be the most important duties to the homemaker? Pick h mending of clothes being a companion to the husband ‘- helping her husband's career — cooking - dish washing ‘- buying of clothes -— entertaining — meal planning : disciplining the children : housekeeping III. Which of the following characteristics would be most desirable in a homemaker? Pick It good-natured progressive _— kind -_ planful —- intelligent -— ingenious -_ modest _ self-controlled : practical : thrifty IV. Which of the following activities should the homemaker consider to be the most important? Pick 1: a helper in her husband's job an earner of outside income to help support the family a keeper of the family social reSponsibilities an organizer of the household a painter, decorator, and beautifier of the home an entertainer of friends and associates a sexual companion to her husband an establisher of a comfortable home a handler of the family finances a companion to her children 109 2 v. v-Ihich of the following attributes should a homemaker have? Pick h __ ambitious ‘ calm __ charming "'- cooperative __ energetic ’- friendly __ good at baking -'— good canner of fruits, vegetables, serves variety of meals "" etc. '7'- __ being informed on new ideas VI. Which of the following activities should the homemaker enjoy performing? Pick ’4 buying foods raising children "‘"" sewing “" socializing with neighbors —— budgeting finances — being interested in husband's work ‘— keeping up own appearance "- decorating the home : cooking : earning own spending money VII. Which of the following qualities would be most advantageous for the homemaker to have? Pick ’4 sense of humor efficient manager careful shopper financially independent (of husband) socially popular enjoys children good cook quick shopper nice looking strict disciplinarian of children VIII. Which of the following duties should the homemaker spend most of her time performing? fick h keeping her husband happy working in local civic enterprises (PTA, Red Cross, etc.) shopping for bargains earning additional family income spending time with the children doing things to help her husband in his job planning meals cleaning the house preparing and serving meals keeping own appearance trim 110 PART III Instructions: Following is a similar list of adjectives or phrases describing behaviors and characteristics of a homemaker. From each group of words or phrases we would like you to select the four which most accurately picture yourself. I. What kind of a home do you run? Pick h fashionable comfortable ‘- orderly -— quiet _- cozy .- modern "'"' efficient "'"" comfortable : relaxed : clean II. What do you feel are your most important duties? Pick )4 cooking being a companion to the husband buying of food entertaining disciplining the children helping your husband's career housekeeping dish washing mending of clothes meal planning . III. Which of the following are most characteristic of you? Pick it kind self—controlled : practical - good-natured intelligent .- progressive "" modest "" thrifty : ingenious : planful IV. Which of the following activities do you consider to be the most important? Pick 1: an entertainer of friends and associates an earner of outside income to help support the family a keeper of the family social responsibilities a handler of the family finances a helper in your husband's job a sexual companion to your husband an establisher of a comfortable home a painter, decorator, and beautifier of the home a companion to your children an organizer of the household 111 2 V. ‘fihich of the following attributes do you have? Pick h cooperative being informed on new ideas __ friendly serves variety of meals ___ calm :::'good canner of fruits, vegetables, .___ ambitious etc. _‘__ charming energetic VI 0 aniCh 0f : good at baking the following activities do you enjoy performing? Pick h cooking keeping up own appearance socializing with neighbors raising children budgeting finances earning own spending money decorating the home buying foods sewing being interested in husband's work VII. ”Which of the following qualities are your greatest asset. Pick h efficient manager nice looking quick shopper strict disciplinarian with children good cook financially independent (of husband) enjoys children careful shopper socially popular sense of humor VIII. 'Which of the fbllowing duties do you spend most of your time performing? Pick 4 keeping your own appearance trhn keeping your husband happy cleaning the house working in local civic enterprises (PTA, Red Cross, etc.) shopping for bargains Spending time with the children preparing and serving meals earning additional family insane planning meals doing things to help your husband in his job. hm.._ - w,— Jain-A Pod/ft has / 51.1.24 --._..}",{-‘E __.—, on“ .. _. _