AGGREGATE DEPARTMENT STORE IMAGES: soGIAE ANE :ExPERIENIIAL FACTORS .: ThBSIS 1or’theDegree G)f F11... D; g. MICHIGAN sIAIE IINIIEIIsIII . . EoEEIII GEORGE NIGIIIIANI ; V . 1967 ‘ WWfi— fHESlS 0-169 . 2 IIININIINIIIIIIII , A A E Michigan State University This is to certify that the thesis entitled Aggregate Department Store Images; _7.- .2 Social and Experiential Factbrs presented by Robert George Wyckham - has been accepted towards fulfillment of the requirements for m degree inWW q“ fl Da'te ML ABSTRACT AGGREGATE DEPARTMENT STORE IMAGES: SOCIAL AND EXPERIENTIAL FACTORS by Robert George Wyckham The importance of the department store image, as perceived by consumers, has grown with the changes in consumer affluence and buying patterns. This thesis seeks to investigate the affects of social influence and various types of shopping experience on consumers' images of department stores. As a theoretical basis for the research a survey of the literature of perception, motivation, interpersonal response traits, attitudes and the societal influences on behavior was made. In addition, a synthesis of the literature on the classical and modern uses of the term image was developed. Data were gathered by means of personal interviews with a randomly selected sample of male and female heads of households from the Detroit area population. An adaptation of Osgood's Semantic Differential was used to measure the direction and intensity of respondent atti- tudes toward various aspects of three test department stores and a hypothetical "ideal" department store. Robert George Wyckham Subjects were classified by social class, using Warner's Index of Social Characteristics, and by various other demographic characteristics and shopping practices. The results of the study show that consumers do have differentiable images of particular department stores. In addition it is evident that social class and race condition the images consumers hold of department stores. However, family life cycle stage and the sex of the subject do not affect the consumer's image of a de— partment store. Attitudes toward shopping have a definite influence on the images people hold of department stores. But shopping companionship practices have little affect. Loyalty to a department store, reading that store's ad- vertisements and social support for beliefs about that store result in favorable and differentiable images of the store. Other shOpping practices such as shopping recency, method of payment and shopping location within the store do not affect the consumer's image of a depart— ment store. AGGREGATE DEPARTMENT STORE IMAGES: SOCIAL AND EXPERIENTIAL FACTORS By Robert George Wyckham A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation 1967 . V1— 647/4349? {.10 W © Copyright by ROBERT GEORGE WYCKHAM 1968 ACKNOWLEDGMENTS A doctoral dissertation is never the work of only one individual. I would like to express my thanks to those who assisted me in this study. Dr. William Lazer, Professor, Department of Market— ing and Transportation, who, as chairman of my dissertation committee, invested countless hours in the completion of this work. Dr. W. J. E. Crissy, Associate Dean, Graduate School of Business Administration, whose ideas and enthusiasm were a constant source of encourage- ment. Dr. Stanley C. Hollander, Professor, Department of Marketing and Transportation, who, as an expert in retailing, provided a realistic approach to the research. The J. L. Hudson Company, Detroit, Michigan, whose financial assistance and guidance made this study possible. My thanks to Mr. James Molineux for his assistance in pPOgramming and statistical analysis. Finally, my deep gratitude to my wife, Judy, for her help, understanding and patience. iii TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES . . LIST OF FIGURES . . LIST OF APPENDICES . . . . . . Chapter I. II. III. INTRODUCTION Background The Problem . . Framework of the Study Method of Research. . Limitations of the Study. Some Possible Contributions of the Study to Marketing Theory and Practice . Organization of the Study FACTORS IN THE IMAGINAL PROCESS Factors Affecting Attitude Formation. Attitudes. . . Images. . IMAGERY AND MARKETING. Brand Images. Self Image . Corporate Image. Department Store Image Summary . iv Page iii vi ix Chapter Page IV. RESEARCH DESIGN: SAMPLE AND QUESTION— NAIRE O O O O O O O O O I C 72 Sample Design . . . . . . . 72 Sample Selection Method . . . . . . 74 The Questionnaire . . . . . . . . 79 Demographic Questions. . . . . . . 88 Field Work . . . . . . . . . . 95 Sample Responses . . . . . . . . 96 IV. DEPARTMENT STORE IMAGE: AGGREGATE IMAGES OF TEST AND "IDEAL" STORES . 105 Method of Analysis. . . . 105 Possible Factors Affecting Aggregate Department Store Images . 107 Aggregate Images of Test and "Ideal" Department Stores . . . . . . 108 “VI. DEPARTMENT STORE IMAGE: SOCIAL CLASS AND FAMILY LIFE CYCLE FACTORS . 124 Social Class Affect . . . . . . . 124 Family Life Cycle Affect. . . 131 138 Social Class—~Family Life Cycle Affect . *VII. DEPARTMENT STORE IMAGES: SELECTED DEMOGRAPHIC CHARACTERISTICS AND SHOPPING . . 148 HABITS. . . . . . . . . . The Affect of Race and Sex on Department Store Image . . 1A8 The Affect of Shopping Attitudes and Shopping Companionship Practices on Department Store Image . . . 154 The Affect of Consumer Shopping Practices on the Image of Hudson's . 160 175 Conclusions on Hypotheses . . . 175 Implications of the Findings for Marketing Research and Strategy . . 184 VIII. SUMMARY AND CONCLUSIONS Some Additional Questions . . . 190 BIBLIOGRAPHY . . . . . . . . . 192 213 APPENDICES. Table 10. ll. 12. LIST OF TABLES Department Store Image . . Scales for Making Primary Ratings of the Status Characteristics of the I. Social-class Equivalents for I. S. C. Ratings . . . . . . . . Social Class of Respondents of Usable Sample Responses . . Family Life Cycle Stage of Usable Sample Responses . . . . . Race of Usable Sample Responses Sex of Respondents of Usable Sample Responses . . . . . . Age of Respondents of Usable Sample Responses . . Education of Head of Household of Usable Sample Responses . . . . . Household Income of Usable Sample Responses . . . Summary of Wilcoxon T and Probabilities Associated with the Null Hypotheses by Social Classes for Image Dimensions Be— tween Test and "Ideal" Department Stores Summary of Wilcoxon T and Probabilities Associated with the Null Hypotheses by Family Life Cycle Stages for Image Dimensions Between Test and "Ideal" De- partment Stores S. C.. Page 64 92 93 99 100 101 101 102 103 103 109 113 Table Page 13. Summary of Wilcoxon T and Probabilities Associated with the Null Hypotheses by Family Life Cycle Stages Within Social Classes for Image Dimensions Between Test and "Ideal" Department Stores . . . 116 Summary of Mann-Whitney U and Probabilities Associated with the Null Hypotheses for Image Dimensions of Test and "Ideal" Department Stores Between Social Classes . 125 14. 15. Summary of Mann—Whitney U and Probabilities Associated with the Null Hypotheses for Image Dimensions of Test and "Ideal" Department Stores Between Family Life . . . . . . . . 132 Cycle Stages Summary of Mann—Whitney U and Probabilities Associated with the Null Hypotheses Be— tween Family Life Cycle Stages Within Social Classes for Image Dimensions of Test and "Ideal" Department Stores . . . 139 16. 17. Summary of Mann-Whitney U and Probabilities Associated with Null Hypotheses Between Races for Image Elements of Test and . . . . . 149 "Ideal" Stores. 18. Summary of Mann—Whitney U and Probabilities Associated with the Null Hypotheses Be- tween Sexes for Image Dimensions of Test and "Ideal” Stores . . . . 152 19. Summary of Mann—Whitney U and Probabilities Associated with the Null Hypotheses Be— tween Consumers Classified by Shopping Enjoyment for Image Dimensions of Test and "Ideal" Stores . . . . . . 155 20. Summary of Mann-Whitney U and Probabilities Associated with the Null Hypotheses Be— tween Consumers Classified as to Shopping Companionship for Image Dimensions of Test and "Ideal" Stores. 158 21. Summary of Mann—Whitney U and Probabilities Associated with the Null Hypotheses Be— tween Consumers Classified as to Shopping Recency for Image Dimensions for a Test . . 161 Store. Table 22. 23. 2A. 25. 26. Page Summary of Mann-Whitney U and Probabilities Associated with the Null Hypotheses Be- tween Consumers Classified as to Shopping Loyalty to a Test Store for Image Dimensions. . . . . . . . . . 163 Summary of Mann-Whitney U and Probabilities Associated with the Null Hypotheses Be- tween Consumers Classified as to Their Method of Payment to a Test Store for Image Dimensions. . . . . . . . . 166 Summary of Mann-Whitney U and Probabilities Associated with the Null Hypotheses Be- tween Consumers Classified as to Shopping Location Within a Test Store for Image Dimensions. . . . . . . . . . . 168 Summary of Mann—Whitney U and Probabilities Associated with the Null Hypotheses Be— tween Consumers Classified as to Reader- ship of Test Store' 5 Advertising for Image Dimensions. . . . . . . . 170 Summary of Mann-Whitney U and Probabilities Associated with Null Hypotheses Between Consumers Classified as to Social Support for Their Beliefs About a Test Store for Image Dimensions. . . . . . . . . 173 viii LIST OF FIGURES Figure Page 1. Processes of Attitude Formation . . . . 18 2. Image Development. . . . . . . . . A6 ix Appendix A. B. LIST OF APPENDICES Household Size Presence of Children Household Income Household Tenure, Value, Monthly Rent Questionnaire Instructions for Interviewers Image Dimensions Statistical Procedures Page 214 216 218 220 222 239 2A7 251 CHAPTER I INTRODUCTION Background Department store executives have long been con— cerned with the consumers' images of their stores. In the first quarter of the twentieth century numerous .articles described the importance of a department store's personality, character or individuality. It was held as a basic principle that to be successful a department store must have a clear, definite image. In 1927 Kenneth Collins wrote that each department store has "An individuality that distinguishes it from its competitors. The problem is . . . to get its individuality across to some portion of the buying public."l Putter2 advocated an overall corporate policy encompassing mer— Chandise selection, delineation of the store's clientele and purchase motivations as an expression of the store's lKenneth Collins, "Institutional Advertising," JOurnal of Retailing, III (April, 1927), p. 10. 2Vita S. Putter, "Store Policy and Personality," M of Retailing, XIV (December, 1938), p. 108. personality. Hotchkin3 and EdwardsLl argued that the department store should use institutional advertising to build the reputation of the store, dramatize its position in the community and build confidence in its merchandise and services. Following World War 11 changes in consumers buying behavior, the growth of suburban retailing and the rise of the discount store resulted in increased managerial concern regarding the image of the department store. Increases in disposable income and education have made consumers more sophisticated, more discriminating and more demanding.5 Greater mobility, within and between urban centers, has increased the size of the consumers' shopping area.6 Suburban retailing and the discount store have caused changes in consumers' buying patterns and vice 3W. R. Hotchkin, "The Present Trend in Advertising," Journal of Retailing, II, p. 5. “Charles M. Edwards, Jr., and W. H. Howard, Retail Advertising_and Sales Promotign_(New York: Prentlce Hall, 1936), p. 162. 5Pierre Martineau, "The Changing American Consumer," Mérketing in Action: Readings, William J. Shultz and Edward M. MazZe, editors (Belmont, California: Wadsworth Publishing Company, 1963), p. 36; Stuart U. Rich, ShOpping Eéhavior of Department Store Customer§_(Boston: Division Of Research, Graduate Sch601 of Business Administration, Harvard University, 1963), p. 1; Howard Rosenborough,. "Sociological Dimensions of Consumer Spending," Canadian IQurnal of Economics and Political Science, XXVI (August, 1960),Ip. 452. 6Rich, Ibid. versa. One strategy employed by department store execu- tives to counter these challenges has been the investi— gation, analysis and attempted re—creation of store images. The importance of the image concept to store manage— ment is inherent in the process of consumer perception. Consumers perceive products and retail institutions not only as physical objects, but as,"psychologica1 things, as symbols of social patterns and strivings."7 The department store is seen as more than its physical plant, salespeople, goods and services. There is, in addition, a psychological nature known as its personality, repu— tation or image.8 Although management policy and action may affect the image of a department store, the image is not the property of the store, but the property of the individuals perceiving the store. The image results from eXperiences with the store. The image individuals hold of a department store determines how they perceive that store and thus affects 9 their shopping behavior. Information about a department M 7Sidney J. Levy, "Symbols By Which We Buy," Advanc- ing Marketing Efficiency, L. H. Stockman, editor (Chicago: Proceedings of the Conference of the American Marketing Association, December, 1958), p. “10. 8Pierre Martineau, "The Personality of the Depart— ment Store," Harvard Business Review, XXIII (January— February, 1958), p0 A7; R. H. Myers, "Sharpening Your Stoie Image," Journal of Retailing, XXXVI (Fall, 1960), p. 29. 9Martineau, Ibid., p. A7; Stuart U. Rich and B. D. store's image held by the total consumer population, and groups within it, is necessary for the development and implementation of sound corporate policy. The Concept of the Department Store Image For the purposes of this study, aggregate department store image is defined as the summation of consumers' images of that store. An individual's image of a depart— ment store is defined as the summation of all of a per— son's attitudes which result from personal experiences, actual, imagined or vicarious, with various facets of that store. Image is a simple construct for a complex inter— action of attitudes resulting from the tendency of the human mind to classify and abstract. The concept of image is important in understanding consumer behavior because it indicates that it is not only external facts and infor- mation which determine behavior. It is not what is true, but what is believed to be true that governs behavior.10 The images individuals hold of particular department stores are affected among other things, by their social relationships. Individuals develop and learn within a social context which has a profound affect on how they Portis, "The Imageries of Department Stores," Harvard W, XXVIII (April, 19611), p. 10. __._.___ 10Kenneth E. Boulding, The Ima e (Ann Arbor: The University of Michigan Press, 195 , p. 7. interpret experience. They tend to adopt a value and belief structure similar to that held by members of their social and economic reference groups.11 It follows, therefore, that various groups of indi— viduals would have different images of particular depart— ment stores because of different values, expectations and desires. These differences are amplified because indivi— duals tend to be exposed to external facts, and perceive and remember them in a selective fashion according to a preconceived set of attitudes. The Problem Statement of the Problem This investigation is concerned with the effect of social influence and various types of experience on the images of department stores held by consumers. Particu- larly, it is concerned with the degree of communality and differentiability of department store images held by con— sumers who are members of groups classified by social and experiential factors. This study seeks answers to the following questions. Are differentiable department store images held by members of: llFrancis S. Bourne, Group Influence in Marketing and Public Relations (Ann Arbor: Foundation for Research On Human Behavior, 1956). 1. different social classes, 2. groups at different stages of the family life cycle, 3. groups which differ in their shopping loyalty to a particular department store, A. groups which differ in their shopping practices with a particular department store, 5. groups which differ in their attitude toward the act of shopping, 6. different races, 7. different sexes? Framework of the Study A survey of the literature of perception, motivation, interpersonal response traits, attitudes and the societal influences on behavior, is presented to provide a back— ground for the research. A synthesis of the literature on the classical and modern uses of the term image is outlined to set the stage for the operational definition of depart- ment store image adOpted. Qperational Definitions of Igrms Used Aggregate department store image is the summation of consumer's images of a department store. A consumer's image of a department sto§e_is the Summation of a consumer's attitudes toward a department store. Attitudes are inferred states of readiness to re— act in an evaluative way toward an object in a situation. Opinions are verbalizations of attitudes.12 An aggregate differential department store image is an image which contains attitudes of a direction and inten- sity which distinguishes it from other images. Communality of aggregate department store images is the tendency for a collection of images to contain atti— tudes which are similar in direction and intensity. Image elements are the attitudes which make up the image of a department store. For example, a consumer's attitude about the courtesy of the sales personnel of a department store is an element of his image of that store. Image dimensions are combinations of elements which center on particular aspects of a department store. A consumer's attitudes about all the facets of the sales personnel of a store make up a dimension of his image. Social classes are, "Groups of peOple who are more or less equal to one another in prestige and community status:" and who tend to, "Share the same goals and ways of looking at life."13 12L. L. Thurstone and E. J. Chave, The Measurement 9£_Attitude (Chicago: University of Chicago Press, 1929), p. 70 1. 13Richard P. Coleman, "The Significance of Social Stratification in Selling," Proceedings of the 43rd National Conference of the American Marketing Association, Martin L. Bell, editor (December, 1960), pp. 157-158. Family life cycle stages are discrete units of time in a family's existence described in terms of marital status, age, and the presence of children. 1” Hypotheses The following are the hypotheses to be tested in this study. 1. 1A Differential aggregate images of particular department stores are held by consumers. Differentiable aggregate images of particular department stores are held by members of different social classes. Differentiable aggregate images of particular department stores are held by members of groups which are at different stages of the family life cycle. Differentiable aggregate images of particular department stores are held by members of social classes which are at different stages in the I family life cycle. Differentiable aggregate images of particular department stores are held by members of groups which differ in their attitude toward shopping. John B. Lansing and Leslie Kish, "Family Life Cycle As An Independent Variable," American Sociological Review, XXII (October, 1957), 512-519- 6. Differentiable aggregate images of particular department stores are held by members of groups which differ in their shopping companion- ship practices. 7. Differentiable aggregate images of particular department stores are held by members of different races. 8. Differentiable aggregate images of particular department stores are held by members of different sexes. 9. Differentiable aggregate images of a particular department store are held by members of groups which differ in their shopping practices with that'store. 10. Differentiable aggregate images of a particular department store are held by members of groups which differ in the social support for their beliefs about Hudson's. Method of Research Data for testing the above hypotheses were obtained means of personal intervieWS of a random sample of 1e and female household heads drawn from the Detroit ea DOPUIation. The random, stratified, multi—stage, ea sample used was selected from the Detroit Standard 10 15 A sample size of 720 atropolitan Statistical Area. is chosen so that no subsample would include less than nirty observations. Chapter IV contains a detailed escription of the sample design. An adaptation of Osgood's Semantic Differentiall6 as designed to measure the direction and intensity of espondent attitudes towards various aspects of three est department stores, a hypothetical "ideal" depart— ent store, and the activity of shopping. The test stores re: the J. L. Hudson Company, Sears Roebuck and Company nd Federal Department Stores, Incorporated. Classification of respondents into social classes as accomplished by means of an adapted version of Warner's gdex of Social Characteristics.l7 Respondents were lassified into stages of the family life cycle in a mnner similar to that used by Lansing and Kish.l8 .— ._‘ 15U. 8. Bureau of Census, U. S. Census of Popu- gtion and Housing: 1960; Census Tracts, Final Report EE_(l)-AO (Washington: U. S. Government Printing Office .962). 16L. E. Osgood, G. J. Suci, P. H. Tannenbaum, The @asurement of Meaning (Urbana: University of Illinois Tess, 1957). 3 l7w. Lloyd Warner, Social Class In America: The flgluation of Status (New York: Harper and Row Pub- ishers, 1960), Chapters 8—15. l8Lansing, op. cit., p. 513. 11 LimitationS'of'the‘Study The results of this study may preclude generali- zation because of the following factors. 1. The information from the sample was obtained at only one point in time. 2. The study was limited to one city. 3. The study was carried out in a city where one traditional department store dominates the market. A. The study deals only with images and they are not the only factors which affect consumer behavior. Some Possib1e_Contributions of the Study to Marketing Theory and Practice In the marketing literature, numerous papers discuss the effect of social class on consumer buying practices.19 lgsee for example Kurt Mayer, "Diminishing Class Differentials in the United States," Marketing and the Ephavioral Sciences, Perry Bliss, editor (Boston: Allyn and Bacon, Publishers, 1963), pp. 185-207; Burleigh B. Gardner, "Behavioral Sciences As Related to Image Build- ing," New Directions in Marketing, F. S. Webster, editor (Chicago: 'Proceedings of the Conference of the American Marketing Association, June, 1965), pp. INS-150; Pierre Martineau, "Social Class and Spending Behavior," Journal 91: Marketing, XXIII (October, 1958), 121-130; Lee Rain- water, Richard Coleman and Gerald Handel, Workingman's Wife (New York: Oceana Publications, 19597? Margaret C. Piere, "Marketing and Social Class: An Anthropologist's View," The Management Review, XLIX (September, 1960): pp. 45-A8; Women and Department Store Newspaper Adver— tising (Chicago: Social Research, Inc°9 195733 Charles J. Collazzo, Consumer Attitudes and Frustrations in Shopping (New York: National Retail Merchants Associ- ation, 1963). 12 These articles assume differences in perception and atti- tudes among consumers in different social classes. Very little empirical data are presented to support to refute the existence and nature of this phenomenon. If these data do exist they are not in the public domain. This study will add to the available knowledge of the influ— ence of social class on consumer perception and attitudes. This study attempts to discover whether persons in different stages of the family life cycle hold different images of particular department stores. It is unlike previous marketing research using the family life cycle concept which related stage in the life cycle to spending behavior.20 This investigation yields information regarding images of department stores held by members of each social class at various stages of the family life cycle. Thus, it allows comparison between the aggregate images of a department store held by older childless couples in various social classes. This is a unique contribution Of this study. The study makes possible an examination of the differences in consumer perception of department stores 20Lansing, op. cit.; J. B. Lansing and J. M. Morgan, "Consumer Finances Over the Life Cycle," Consumer Behavior Vol. II, Lincoln H. Clark, editor (New York: New York University Press, 1955), pp. 36—51; S. G. Barton, "The Life Cycle and Buying Patterns," Consumer Behavior, Vol. II, Lincoln H. Clark, editor (New York: New YorETUni— versity Press, 1955), pp. 53-57. 13 .mong consumers classified by various shopping experi— Inces and practices. ShOpping habits examined are: shopping companionship practices, store loyalty, shopping 5ecency, shopping location within store, payment method 1nd advertising readership. In addition, the effect of :onsumers' attitudes toward shopping and the social sup— )ort for a subjects' beliefs about a department store on iepartment store images will be analyzed. The research facilitates the comparison of the images of the three department stores studied. Compari— ons among the test stores and the "ideal" store will allow examination of each store's perceived strengths and weaknesses. An individual department store may then be in a better position to make corrections in its weaker areas and exploit its strengths. Store executives may also be able to formulate policies more in line with the con- sumers' "ideal." By breaking down the total image profile of a depart— ment store into the image profiles h61d by various groups of consumers additional information for policy determi— nation may be obtained. Consumers, classified by demo— graphic and experiential data, comprise market segments. Phe images held by these market segments may suggest corporate strategy in merchandising, pricing, sales per— sonnel and advertising. 14 Organization‘of'the'Study The study is divided into eight chapters. Chapter I esents the background of the problem and the concept of m department store image. The problem is outlined, re- :arch questions posed, terms defined and hypotheses pre— ented. Research methodology is described as are the imitations of the study. Possible contributions of the tudy to marketing theory and practice are reviewed. Chapter II outlines the functioning of perception, otivation and interpersonal response traits in the reation and operation of images. The interrelationship f these three psychological processes is analyzed and he affect of the psychological and sociological environ— ent on perception, motivation and interpersonal response raits is examined. Theories of attitudes, attitude for— btion and attitude measurement are outlined. Societal lnfluences are discussed in terms of group membership, tatus and role, the family life cycle and social class. Chapter III presents an historical outline of the ,oncept of image taken from the psychological, sociological Ind marketing literature. Contributions of the classical :oncept of image to the modern concept of image are re— Viewed and the uses of the term image in modern social science and marketing are outlined. The relationship of rarious marketing images: SGlf image and consumer be- lavior is discussed. l5 The research design and sample responses are pre— nted in Chapter IV. Sampling considerations, sample pe and size and the sample selection procedure are ascribed. The questionnaire is outlined under the )llowing headings: the semantic differential, selection f test department stores, the activity of shopping, de— artment store shopping and demographic questions. Field ork is discussed in terms of selection, training and ompensation of interviewers. The sample responses, actors affecting sample composition and selected demo- raphic characteristics of the sample are reviewed. Chapter V is comprised of an analysis of the survey esponses on department store images. Some factors which my have affected respondents' attitudes toward the test mores are described and the method of analysis is pre- I ented. The images of the test stores and the "ideal" tore are compared. The images of the test and "ideal" tores held by consumers in different social classes, in ifferent family life cycle stages, and in each social lass at various stages of the family life cycle. Chapter VI presents the affect of social class, amily life cycle and social class at various life cycle tages on department store images. Marketing research nd strategy implications of the results are discussed. Chapter VII gives additional findings on the effect fi‘demographic characteristics and shopping habits on 16 Ipartment store image. The affect of race, sex, shopping :titudes and shOpping companionship practices on the nages of the test and ideal stores is presented. The ffect of shopping practices and social support for be— iefs about Hudson's on the image of Hudson's are nalyzed. Chapter VIII presents the summary and conclusions I the study. The objectives and hypotheses of the study we summarized and conclusions regarding the hypotheses Ire presented. Implications of the findings for market- Lng research and strategy and some additional questions )f practical and theoretical importance are discussed. CHAPTER II FACTORS IN THE IMAGINAL PROCESS This chapter has two objectives. The first is to escribe the factors which influence attitude formation. The second is to outline the development of images, which we composed of attitudes. These objectives will be accomplished by a survey of the literature of perception, motivation, interpersonal response traits and group theory as they are related to attitudes and images. The litera- ture review in this chapter will serve as a basis for the discussion of marketing and imagery in the next chapter. Factors Affecting Attitude Formation Figure 1 depicts the process of attitude formation. ttitudes are formed within and are affected by the physi- cal, psychological and sociological environments. Within these environments the individual's perception, motivation and interpersonal response traits interact in creating an attitude toward an attitude object. The resulting attitude, in turn, affects the person's perception, motivation and interpersonal response traits. l7 Ioaoanopcfl mflnp QwSOQQB on» mo pxopcOo can snap no QHQmQOHpmHonnopeH mpsomo ram... 23:2. u .mpsoesonfl>zo Hmowmoao H3 mononp.omeommon Hos neon anemone .ooenon one mooSBflppw QHQmQOHp Hoom one HmonOHonommo «Hmofiwoaowwmno omnoonopefl one soapm>fipoe neonpaoonom “soapwenom oUSBpro no mommooonmll.a maswflm Inc—concise...“ .325 umzommuc 18 U0 arr—1.. < aczowmmccusz. ‘Illll I humane 22.33: we I _ moat»: 20.523: GIL pzuzzoczéu 1330.3: 0 3.. L l9 pppeption "The first stage in attitude formation——in the cst complicated social situation as well as in the re— tricted laboratory experiment—~is the perceptual tage."l To better understand attitudes and thus images he perceptual process and the principles of perception ill be discussed. The parceptual process.-—Perception is the process V which a person structures the raw data he receives .hrough his sensory organs. It is through this act of mganization that the individual gives meaning to ob— ects, other persons and situations. Historically per— eption was conceived of as a phenomenon of the conscious Iind.2 Modern thought emphasizes the unconscious and mn-verbal as well as the conscious and verbal nature of 3 erception. Sherif and Cantril state that the determinants of erception result in, ". . . referential frameworks and that these . . . frames serve as anchorages to A tructure or modify subsequent eXperience and response." ng—L lM. Sherif, "A Study of Some Social Factors in erception," Archives of Psychology, July, 1935, p. 327. 2J. R. Kantor, "Suggestions Towards a Scientific nterpretation of Perception," Psychological Review, XVII (1920), 191—216. 3Pierre Martineau, "It's Time to Research the Con— g§§§,"uHarvard Business Review, XXXIII (July—AUBUSt) 6 3 o A Muzafer Sherif and Hadley Cantril, "Psychology of 20 An individual's perception of an object or a situation is a product of: (1) his physical and social environ— ments, (2) his physiological structure, (3) his wants and goals, (A) his past experiences,5 (5) the purpose of the perception, and (6) the stimulus forcing action. Principles of perception.-—7The direct relation— ship between perception and attitudes necessitates an awareness of the principles of perception. For a clearer understanding of the formation of attitudes we shall dis— cuss the principles of perception resulting from the work of Gestall psychologists as well as the contributions of some of the more sociologically oriented students of per- ception.8 Attitudes, Part II," Psychological Review, LIII (January, 19A6>, 19-20. 5David Krech and Richard S. Crutchfield, Theory and Problems of Social Psychology (New York: McGraw— Hill Book Company, Inc., 19A8), p. 30. 6Hadley Cantril, "The Nature of Social Perception," Imman Behavior from the Tgansactional Point of View, F. P. Kilpatrick, editor IHanover, New Hampshire: Institute for Associated Research, 1953), p- 225- 7Most of the material for this section has been ob— tained from G. W. Allport, "Attitudes," A Handbook of §gcia1 Psychology, Carl Murchison, editor (Worchester, Massachusetts: Clark University Press, 1935), p. 836; Bernard Berlson and Gary A. Steiner, Human Behavior: An Igyentory of Scientific Findings (New York: Harcourt Brace and World, Inc., 196AI, Chapter IV; David Krech, Richard S. Crutchfield and Egerton L. Ballachy, Individual IE_Society (New York: McGraw—Hill Book Company, Inc., 1962). p167 77-89. See for example, Michael Wertheimer, Readings in GEEEEQEIQE (Princeton: D. Van Nostrant Company, Inc., 21 The principle of configuration is that perceives aggregates not as aggregates but as unified wholes.9 A sub-principle of configuration is Closure—-the ability to perceive a portion of an aggregate as a whole. Objects are perceived against a background. This is known as the principle of figure and ground. EXperience with an object or situation leads to an interpretation which is carried forward to later per— ception. Thus an identical or similar object will be perceived in light of prior knowledge. Scott, at the turn of this century, called this phenomenon‘app'e'r'c‘epti'on.lO Similarly, individuals develop frames of reference or anchorages from past experience. These are used as the basis of judgment for new eXperiences.ll Cantril describes the learning effect of perception and the subsequent frames of reference as surety of_perception. 1958); M. D. Vernon, The Psychology of Perception (Balti- more: Penguin Books, Inc., 1962); Floyd H. Allport, Theories of Perception and the Concept of Structure (New York: John Wiley and Sons, Inc., 1955), Chapter V.- 9E. B. Titchener, A Beginner's Psychology (New York: Macmillan, 1915), p. 115. lOW. D. Scott, The Theory of Advertising (Boston: Small Maynard and Company, 19097, pp- 199-150. llKrech, op. cit., p. 32. 22 We experience surety in our perceptions if hey have in the past proved to be reliable guides o purposive action. . . . Surety of perception . . . is . . . reflected in the speed and con— istency of judgment and action.1 he perceptual principle of gap states that per- is governed to some extent by what an individual y to perceive. Allport says we have, "perceptual ncy," hypotheses from past experience that tell _objects to look for and, to some extent, how 13 bjects will be likely to appear. All of the 1, social and personal factors which have gone individual's beliefs, values, wants, attitudes pectations are part of his set. There is communality as well as differentiability ceptual phenomena. Common elements in eXperience ad to some form of commoness in perception. This city of perception allows the development of ate or public images of people, objects and situ- Influences of the social environment ongperception.-_ born, develops and exists within a social environ— )ich affects perception. In recent years consider— Iterest has been displayed in the affect of this '2Cantril, op. cit., pp. 225-226. ‘3Allport, op. cit., pp. 381. 23 °onment on the invidual's perceptual processes. rch by psychologists, economists, marketers, and sociologists has shown the impact of primary groups 15 rception. Secondary groups such as social organ— ons, political parties,16 educational institutions 17,18 ocial classes have been found to influence per— Ion. In addition, perception is affected by statistical )s classified by age, sex, education, income and family cycle stage.19 1”See for example, E. A. Rogers, The Diffusion of rations (Glencoe, Illinois: The Free Press, 1960); 1 L. Child, "Socialization," Handbook of Social Eology, Lindzey Gardner, editor (Cambridge, Massachu— Addison-Wesley Publishing Company, Inc., 195A); Katz and Paul F. Lazarsfeld, Personal Influence coe, Illinois: The Free Press, 1955). 15See for example, Rogers, op. cit.; Elizabeth Ist, "Do Husbands or Wives Make the Purchasing Iions"? Journal of Marketing, XXIII (October, 1958), -58; William H. Whyte, Jr., The Organization Man York: Simon and Schuster, 1956); Katona, op. cit. 16Katz and Lazarsfeld, op. cit. l7Pierre Martineau, "Social Class and Spending ior," Jgurnal of_Marketing, XXIII (October, 1958), 30. 18 Tomotsu Shibutani, "Reference Groups as Per— ives," Marketing and the Behavioral Sciences, Bliss, editor (Boston: Allyn and Bacon, Inc., 1 98cc for example, Henry L. Munn, "Brand Perception .ated to Age Income and Education," Journal of 331g, XXIV (January, 1960), 29—324. 2A otivation is the second of the interacting ogical processes which influence the creation of es. Man's behavior is directed by what he per— his world to be. Why man acts in a particular depends on his motivation. The problem of motivation is taken to mean he problem of explaining behavior, and a motive 5 often thought of as anything that moves the rganism; that is to say, anything that affects ehavior.20 odern motivation theory.——Modern motivational is based primarily on drive theory. "This con— s introduced by Robert S. Woodworth in 1918 to e the energy that impels an organism to action as d to the habits that steer behavior in one direction ther."21 Cannon's concept of homeostasis led to a nition of the drive theory of motivation as the resulting from homeostatic imbalance or tension. Aaslow's concept of motivation is primarily a drive He suggests that there is a hierarchy of motives is. First are the physiological needs, the safety the belongingness and love needs, the esteem needs 0K. F. Walker, "The Nature and Explanation of r," Psychological Review, XLIX (19A2), 581. 1 E. J. Murray, Motivation and Education (Engle- iffs, New Jersey: Prentice-Hall, Inc., 196A), 25 22 As lower inally the need for self—actualization. needs are satisfied higher motives become operative. A definition of motivation.--A1though concensus theorists on a definition of motivation is not ble, it is clear that motivation is both internal xternal. Needs are internal factors that arouse, 23 t, and integrate a person's behavior. Incentives, als, are external and influence the direction of a n's behavior. Man's needs and wants are interactive with his es. Motivation includes both positive and negative ng forces. Wants and needs are positive forces which a person toward certain objects or situations. 3 and aversions are negative forces which repel a )n from objects and situations.24 Both of these forces .nitiating and sustaining forces of behavior. Objects 'ds which wants are directed are "approach objects;" ts which repel are "avoidance objects." Both approach voidance objects are known as incentives or goals.25 The self and motivation.——The self occupies a Vital in motivation. It is both an organizer and the L— — 22 A. H. Maslow, Motivation ang_Persgnality (New York: e and Brothers, Publishers, 195A), Chapters IV, V. 2 3Murray, 0 . cit., p. 7. 2A Krech, op. cit., p. 69. 25Ibid., p. 66. 26 t of wants and goals related to self enhancement elf defense. Maslow's love and belongingness needs elf actualization needs are interwoven with the n of self.26 Because the self is a product of 1 interaction, self evaluation is mainly a compari- f self with reference groups. An individual's ation of self is primarily a function of the vement of goals reflecting group values.27 ,personal Response Traits Interpersonal response traits are the third of the ological processes important in the formation of udes. They are defined by Krech as enduring tenden— to respond to other people in characteristic ways.28 9 interpersonal reactions are akin to the notion of , but in a different dimension. Both concepts are arily based on social interaction. The concept of , however, describes the individual's actions in )nse to his evaluation of self. The concept of “personal response traits describes the person's :ncies to respond to others. The influence of F 2 6Maslow, 0p. cit., Chapters IV, V. 27 Krech, op. cit., pp. 77-8A. 28 . Ibld., p. 10A. 27 ty, which is not emphasized in discussions of self, eloped in the notion of traits. Interaction of perceptions,4wants and interpersonal se traits.——An individual's perceptions and wants, t, determinp his interpersonal response traits. generalized tendencies to respond to others, in nfluence his perceptions of the world and the in which he seeks to satisfy his wants. Maslow bes this phenomenon in terms of the individual as anized whole. A particular want may change an individual's perceptions, his memories, his emotions, the con— tent of his thinking. This list can be extended to almost every other faculty, capacity or function, both physiological and psychic. Over time the interaction of perceptions, motivations terpersonal response traits become organized into x systems known as attitudes. These attitudes, in are the elements of images. Environment Social structure plays a major role in the particular y of cognitions, wants, response tendencies and des.30 In this section attention will be given to fects of group membership, the family life cycle, cial class on attitudes. 29Maslow, op. cit., pp. 63—6A. 3ONewman, pp, cit., p. 28A. 28 Group membership.——Groups are collections of indi- Ls who share common needs, goals and beliefs. Over the group develops norms of behavior which result and in turn affect the perceptions, motives and 31 ldes of its membership. Every individual belongs rast number of groups not all of which affect atti— greatly. Those groups which significantly affect son's attitudes (those with which he identifies) Iown as reference groups. They may be groups to be belongs or to which he aspires to belong. They ange in size from two members up to large numbers, 1 organization from informal to formal.32 The primary group (family, friends, colleagues) is Lly seen to be the most powerful in controlling its c's attitudes. Next in strength are groups to which adividual belongs and identifies, or just identifies, Kample school, church, social class. The largest east powerful in influencing attitudes are statistical s such as age, sex and race. Much research has been carried out to show the effects Dup influence on perceptions and attitudes. The famous of the autokinetic effect by Muzafer Sherif showed nfluence of the group on the perceptions of the indi- 1. Pairs of subjects were asked to judge a highly l 3 Tomotsu Shibutani, op. cit., p. 28A. 2 3 Bourne, op. cit., pp. 2A7—255. 29 guous stimulus (a visual illusion of movement of a ionary light in a dark room). It was found that the ments of each individual tended to converge toward other.33 Asch's line experiments showed how strong pressure of the group was in altering perception. ,had individual naive subjects, in a group of con— rates of the experimenter, make judgments involving discrimination of lengths of lines. The results cated that naive subjects' judgments tended to move ,rd the erroneous judgments of the confederates of experimenter.BL4 Newcomb's study of female university students' ,tical attitudes showed that group identification and ‘ormity to group norms were instrumental in attitude Ige.35 A study by Sherif of boys in a summer camp .cated attitude change may be, in part, a function of individual's status in the group.36 33Muzafer Sherif, ”A Study of Some Social Factors ’erception," Archives of Psychology, 1935. 3“Solomon Asch, Social Psychology (Englewood ‘fs, New Jersey: Prentice—Hall, 1952), pp. A50-A5l. 35Theodore M. Newcomb, Personality and Social 1E2 (New York: Dryden Press, 19A3). _ 36Muzafer Sherif, ”A Preliminary Experimental Study .ntergroup Relations," Social Psychology at the Cross— EJ John H. Rohrer and Muzafer Sherif, editors (New ' Harper Brothers Publishers, Inc., 1951), p. A08. 30 The family_life cycle.--The family has been de- d as a primary reference group having a large mea- f influence on the perceptions, motives and atti- of the individual. For this reason the family has he subject of research by students of human be- 7. One of the interesting and useful notions re— ng from this research is that of the family life The life cycle concept is based on the thesis: That the changes that occur in people's atti- tudes and behavior as they grow older . . . may be less associated with the biological process of aging than with the influence of age upon the individual's family relationships.37 A family's existence may be thought of in terms of te units of time: the stages in the life cycle. Kample, there is a period when there are no children a family, when children are growing up, when all the ren have left home, and when one Spouse has passed 38 Attitude patterns vary as the family passes gh the stages in the life cycle.39 Social clas§.-~An interrelationship exists between 1 class and primary groups which makes social class I__ 37John B. Lansing and Leslie Kish, "Family Life As An Independent Variable," American Sociolggical E, October, 1957, pp. 512—519. 38For examples of variously defined stages in the Y life cycle see, Lansing, op. cit.; Harold H. Mayer, Adult Cycle," The Annals of the American Academy of lgal and Social Science, September, 1957, pp. 58-67. 39 Gerald Zaltman, editor, Markeping: Contributions ghe Behavioral Sciences (New York: Harcourt Brace orld, Inc., 1965), p. 12. “—_,__.___ -..,1.L LLLL l, 31 important factor in attitude formation than might appear. A social class is composed of primary and, therefore there is an indirect social class nce on the individual through his primary groups. According to Young, "Social class structure con— of a stratified hierarchy of power distributed in "40 .s sub—groups along a preferential scale. In I's terms social classes are: Groups of people who are more or less equal to one another in prestige and community status: they are peOple who readily and regularly inter— act among themselves in both formal and informal ways; they form a 'class' also, to the extent that theZ share the same goals and ways of looking at life. 1 :endency to share common goals and life expectations to a degree of communality of attitudes.42 Newman ts on two studies which support the thesis that social position affects behavior.Z43 uoKimball Young, Hangbqok of Social PsychOlogy n: Ioutledge and Kegan Paul, Ltd}, 1960), Rev. . 22 . ulRichard P. Coleman, "The Significance of Social fication in Selling, Proceedings of the A3rd a1 Conference of the AmeriEan Marketing'ASSOCiatIOn, a L. Bell, editor (December, 1960), pp. 157—158. A 2Shibutani, op. cit., p. 227. H3Newman, op. cit., pp. 256—263, 320-3A2. 32 ';AttitUdES Behavior becomes stable and consistent through the pment of attitudes towards classes of objects, per— and situations. If they (attitudes) did not exist as fairly organized and coherent dispositions in the mental life of_each individual it would be impossible to account for the patent stability and consistency in human conduct. However, attitudes are not the only determinants of ior. Cook and Selltiz list other behavior determi- as: characteristics of the individual; his dispo- , values, and motivational state; and characteristics a situation; the situational norms, the expectations A5 iers and possible consequences of actions. Aptitude defined.—-A11port in an excellent survey a literature, defines attitude and distinguishes it A6 ather forms of response readiness. This definition :itude is eclectic in nature and has withstood the ight of time. An_attitude is a mental and neural state of readiness, organized through eXperience, exerting a directive or dynamic influence upon the indivi— dual's response to all objects and situations with which it is related.” m4Gordon W. Allport, "Eidetic Imagery," British al of Psychology, XV (October, 192A), 99-120. uSStuart W. Cook and Claire Selltiz, "A Multiple ator Approach to Attitude Measurement," Psychologi- alletin, LXII (196A). 37. u6Allport, "Attitudes . . .," op. cit., pp. 798-8AA. u71bid., p. 810. u—‘Afi 33 Forms of response readiness which may be distin- hed from attitudes are: reflexes; conditioned re- es; instincts; habits; needs, wishes and desires; j iments; motor set; interests and subjective values; udices or stereotypes; concepts; opinions; and ts.48 ”Attitude‘Objects.-—The object of an attitude may nything that is within the individual's psychological d. Thomas and Zaniecki argue that: "Since an atti— is always directed toward some object it may be de— d as a state of mind of the individual toward a "49 .e... Attitude components.——An attitude is composed of e interdependent components: the cognitive component, evaluative component and the action tendency com- wnt. The cognitive component is made up of an indi- .al's evaluative beliefs about objects. Emotions con— ed with objects reflect the feeling component and attitudes their motivating character. The action ency component is the behavioral readiness aspect of ttitude.50 u81bid., pp. 806—810. “9w. 1. Thomas and F. Znaniecki, 1918 The Polish ant In Europe and America, Vol. I (Boston: Badger, ’ 1918): p0 8110 T 50 Krech, op. cit., pp. lAO—lAl. 3A Allport describes the same phenomenon, but instead fining components of attitudes he distinguishes two of attitudes. one which is so organized and energized that it actually drives, and the other which merely directs . . . the motivational and the instrumental The manner in which an attitude affects behavior is rt determined by the nature of its components. components may be described in terms of valence and plexity. Valence describes the degree of favorabil— r unfavorability with respect to the object of the 52 ude. Multiplexity refers to the number and kind ements making up the components, i.e., the beliefs, 53 ngs and behavioral tendencies. ring Attitudes 1. Problems of measurement.--The following are some of the problems of measuring attitudes: a. Deception and rationalization by subjects. b. Answers by subjects which purport to be attitudes but which are really only re— 55 sponses to the instrument. Allport, op. cit., pp. 818—819. Ibid., p. 820. 53L. L. Thurstone and E. J. Chave, The Measurement titude (Chicago: University of Chicago Press, 1929), 54Allport, op. cit., p. 836. 55Krech, op. cit., p. 176. 35 c. The "influence of the process of measure— ment upon the subject of measurement ."56 d. The creation of unnatural pictures of attitude structures by forcing them into scales. e. The discrepancy between attitude and pre— dicted behavior and actual behavior.57’58 f. Interaction of attitudes and other factors which influence behavior.59’6o’61 this discussion three important points emerge. One, easurement of attitudes is difficult. Two, it re— s indirect rather than direct methods of measurement. 56Ibid. 57R. T. Lapiere, "Attitudes vs. Actions," Social . XIV (193A), 230-237. 58Thurstone and Chave, op. cit., p. 10. 59G. W. Allport, "The Historical Background of n Social Psychology," Handbook of Social Psychology, er Lindzey, editor (Cambridge, Massachusetts: Addison— y Publishing Company, Inc., 195A), p. 37. . 60B. M. Bass, "Authoritarianism or Acquiescence,” a1 Of Abnormal and Social Psychology, LIV (1957), 3A 61A. L. Edwards, "The Relationship Between the d Desirability of a Trait and the Probability That rait Will Be Endorsed,” Journal of ApplIed Psychology, I (1953), 90—93; A. L. Edwards, The Soc1al Desirabil— ariable in Personal Assessment and Research (New Dryden Press, 1957). _r -d-__ 36 three, attitude measurement is only one factor in )rediction of behavior. 2. Methods of measurement.—-Attitudes cannot be measured directly, but must always be inferred from behavior. Behavior may be a verbal ex— pression, a task oriented activity or a physi— cal reaction to a representative of an object— class.62 Cook and Selltiz' grouping of the techniques of measuring attitudes into five major classifi— cations is similar to Krech, Cruchfield and Ballachey's categories of the measures of motivation.63 a. Measures in which inferences are drawn from self reports of beliefs, feelings and be- havior toward an object or class of ob— jects. For example, inferences may be drawn from a consumer panel's report de- scribing the personality of a department store. b. Measures in which inferences are drawn from observed overt behavior toward an object. For example, inferences may be drawn from 62Krech, op. cit., p. 147- 631bid., pp. 77—89. “as." 37 observation of shopper buying activities associated with competing brands. 0. Measures in which inferences are drawn from reactions to or interpretations of ambiguous material relevant to an object. For example, inferences may be drawn from sentence com— pletion tests regarding proposed advertising symbols. d. Measures in which inferences are drawn from performance of tasks relevant to an object. For example, inferences may be drawn from role playing in which consumers are asked to simulate particular buying practices. e. Measures in which inferences are drawn from physiological reactions to an object.64 For example, inferences may be drawn from the results of teenagers' reactions to new fashion items measured by means of psycho- galvanometer. 3. Attitude scales.—-The most important device used in attitude measurement is the attitude scale. An attitude scale is composed of a set of statements or items to which subjects respond. The pattern of responses leads to inference 6“Stuart W. Cook and Claire Selltiz, "A Multiple dicator Approach to Attitude Measurement," nggpglgg}: 1 Bulletin, LXII (196A), 36—55. 38 about attitudes. The objective of all attitude scales is to place each individual in a numeri- cal position on a continuum. This position in— dicates the direction and intensity of the individual's attitude toward the object in question.65 Among the advantages of attitude scaling are:6 a. The quasi—game situation encourages re- spondent participation. b. Many items may be evaluated quickly thus minimizing fatigue. c. The influence of respondent articulateness is minimized. d. Uniformity of stimulus results in reliability. e. Interviewer bias is minimized. f. Results can be coded and tabulated quickly, easily and objectively. g. Precoding answers insures correct classifi- cation of results. A. Specific measurement instruments.—-A variety of types of attitude measurement devices have been developed. The six principal scaling methods 65Krech, op. cit., p. 1A7. 66See also, Russell I. Haley, "New Insights Into tude Measurement," New Directions in Marketing, F. S. ter, editor (Chicago: American Marketing Association eedings, June, 1965), pp. 309—330- 39 for attitude measurement are: the method of equal appearing intervals developed by Thur— stone and his colleagues;67 the method of summated ratings created by Likert;68 the 69 social distance scale designed by Bogardus; cumulative scaling developed by Guttman;7O the scale—discrimination technique originated by Edwards and Kilpatrick;71 the semantic differ- ential created by Osgood and his associates.72 The semantic differential, the technique to be used in this study, will be described. 73 The semantic differential .——The semantic differ- ntial technique was developed by Osgood, Tannenbaum and uci to measure the meaning of concepts. It is based on 67Thurstone and Chave, op. cit., p. 12. 68R. Likert, "A Technique for the Measurement of titudes," Archives of Psychology, 1932. 69E. S. Bogardus, "Measuring Social Distance," ournal of Applied Sociology, IX (1925), 299— 308. 70L. Guttman, "The Third Component of Scalable ttitudes," International Journal of Opinion and Atti— de, IV (1950), 285—287. 71A. L. Edwards and F. P. Kilpatrick, "A Technique or the Construction of Attitude Scales," Journal of plied Psychology, XXXII (19A8), 37A-38A. 72C. E. Osgood, P. H. Tannenbaum and G. J. Suci, he Measurement of Meaning (Urbana: University of Illinois r988, 1957) A0 3 assumption that an object may have subtle connotative anings as well as obvious denotative meaning. To mea- re the connotative meanings an indirect approach was vised. The subject was asked to indicate the meaning the object to him by rating it on a seven interval ale of bipolar objectives. The meaning of the object r the person is the profile of his ratings on the atti— de scales. Three general factors of meaning were uncovered in factor-analytic study of the ratings of many different jects on bipolar adjective scales. These factors are e evaluative, the potency and the activity factors. The st conspicuous factor is the evaluative factor. It rresponds to the valence of attitude components. The lence of an individual's attitude toward an object may measured by averaging his rating scores on those items t heavily loaded for the evaluative factor. The semantic differential scale with its seven inter- s results in a mid—interval which is neutral. The blems of interpreting the meaning of neutral scores on s scale are important. Does a neutral score indicate "neutral" attitude, an ambivalent attitude, or the lack an attitude? Or does it indicate uncertainty, lack interest or lack of knowledge about the subject? st—retest reliability data for the semantic differential e obtained in one experiment by Tannenbaum. He had 135 Al jects judge six concepts against six evaluative scales two occasions five weeks apart. In this case he nd test-retest coefficients ranging from .87 to .93, h a mean r (computed by z — transformations of .91).?“ The evaluative dimension of the semantic differ- ential exhibits reasonable face—validity as a measure of attitude. Suci was able to differ— entiate between high and low ethnocentrics with the use of the evaluative scales of the differ- ential.75 validity of the semantic differential as a measure of itudes is indicated by the substantial correlations ch have been found between the evaluative ratings and res on Thurstone and Guttman scales. Comparison of ferential ratings and Guttman scores measuring atti— es toward crop rotation yielded a rank order corre— ion of .78 (o .78; p < .01). In a comparison of the antic differential and the Thurstone scale on three cepts, the Negro, the church and capital punishment, as found that the reliabilities of both instruments e high and equivalent. The correlations between the antic differential scores and the corresponding Thur— e scores ranged from r = .7A to r = .82 and was ificantly greater than chance (p < .01) in each case.76 7”C. E. Osgood, op. cit., p. 192. 751bid., p. 193. 76 Ibid., pp. 93—11A. A2 Attitude formation.-—Among the factors important he formation of an individual's attitudes are his Is, the information available to him, his group re— .onships and his personality.77’78 Allport has suggested that there are four processes Vhich attitudes are likely to be formed: (1) inte— ting a number of similar experiences, (2) differ— iating from general to specific situations, (3) un- al experiences, or (A) adopting attitudes from .ers.79’8O In the process of satisfying his wants an individual Ielops attitudes. Any of Allports ”processes" may play >art in the formation of these attitudes. However, the jority of an individual's vital attitudes regarding ne, family, marriage, sex, duty, religion, vocation, ial welfare, politics, etc., are formed in adolescence endure throughout life. "Barring unusual experience conversion or crises, attitudes are likely to be con- med and enriched rather than altered or replaced."81 77 78James Morgan, "A Review of Recent Research on .sumer Behavior," Consumer Behavior, Lincoln H. Clark, Egg (New York: Harper Brothers Publishers, 1958), Krech, op. cit., p. 180. . 79A. Oxenfeldt, D. Miller, A. Shuchman, c. Winick, lghts Into Pricing (Belmont, California: Wadsworth lishing Company, 1961), pp. 87—88. 0 For more complete discussion see, Allport, op. cit., 810—811. 81 Krech, op. cit., pp. 186. A3 tudes become fixed response tendencies for similar cts, individuals or situations. Thus, an attitude serve various goals, and different wants can give (to the same attitude. "Attitudes . . . are also shaped by the infor— "82.83 .on to which a person is exposed Infor— Lon, however, is rarely the determinant of an attitude apt in the context of other attitudes. It is impor— t to note that attitudes do not always reflect reality. fact, certain attitudes develop in men (such as super- tions, delusions and prejudices) which are character— Id by their wide divergence with the facts. Because man is a social being many of his attitudes re their sources and support in the groups to which he ongs and to which he would like to belong. These ial attitudes are intertwined with the group's norms, 8A,85 ues and beliefs. In order to maintain these itudes the individual requires the support of the 86 up. Thus there is an interrelationship between an 82Ibid. . 83See also, James G. March, Herbert A. Simon, anizations (New York: John Wiley and Sons, 1958), 127—12 8A Krech, op. cit., p. 213- 8 5See also, L. L. Bernard, "Attitudes, Social," yclo edia of the Social Sciences, Seligman and Johnson, POTS (New York: Macmillan, 1930), pp. 305—306. 86 . A. H. Hastorf and A. L. Knutson, "The Nature of itude and Opinion,” Human Nature From a Transactional AA idual‘s wants, the information he has or seeks, the s to which he belongs and the attitudes he holds. Ldition, each person tends to accept those attitudes 1 fit into his personality. The personality of the individual . . . is not a perfectly integrated system and the indi— vidual may take over attitudes that are incon— sistent or contradictory because of the different teaching of his authorities in different areas, because of conflicting group affiliations and because of conflicting wants.87 Images Attitudes seldom exist alone but tend to form sters related to an object or group of objects. These sters tend to form total systems known as attitude 88,89 stellations. In this study the term image di- sion will describe an attitude cluster and aggregate e will describe an attitude constellation. At this t it should be clear that the psychological processes h influence the formation of attitudes have a direct ct on the development of images. t of View, F. P. Kilpatrick, editor (Hanover, New shire: Institute for Associated Research, 1953), 23A-235. 87Krech, op. cit., p. 213. 88Ibid., pp. lAA-lAS. 89 See also, Erle Fiske Young, "Balance and Imbalance, Personality," Social Attitudes, Kimball Young, editor York: Henry Holt and Company, 1931), p. 78. A5 Image creation is pictured in Figure 2. The adigm pictures the image of an object as the sum- ion of attitudes towards various dimensions of the cot. This process of image development takes place hin and is affected by the physiological, psycho- ical and sociological environments. 1y Uses of the Image cept Imagery received a great deal of attention, and the subject of much study, during the latter part of 19th century and the early part of the 20th. Images e conceived of as one of the elementary units of think— which was a basic process of the mind.90 Various facets of images were studied. An extended ument over the importance and definition of sensation, 91 ception and image was carried on in the literature. reat deal of experimentation and writing developed out the problem of imageless thought.92 90Henry R. Holt, "Imagery: The Return of the rasized," American Psychologist, XIX (March, 196A), ; Oliver L. Reiser, "The Structure of Thought,” cholo ical Review, XXI (January, 193A), 51-73. 91E. B. Titchener, A Beginners Psychology (New k: Macmillan, 1919), p. 73; Stephen Colvin, "The ure of the Mental Image,” Psychological Review, XV 08), l6A—165. 92Holt, loc. cit.; Robert M. Ogden, "Imageless uEht: Resume and Critique," Psychological Bulletin, I (June, 15, 1911), 18A; Hiram M. Standley, "Language Image," Paychological Review, IV (1897), 69; W. B. page do whommsosflo m50fipm> we cm mo owwsfi one mo pcosqoa flonwwon moozpflppm no so o>oo one mpmomongon Emawmao . O prEE5m one szonnp noonno oompwmmm .p:oEQ0ao>oo owmsHII.m opmmHm changes—z; 411.043.»... :1. I .I . . . 32:. l I . a E 8'. >r_._<...uo:oo . .. .1 . CI... 2.9.: .835. "HI I . - woq z_ 5 All... mu)?» "30.... < . .E. lllllllll . . All 5531:.» lUIhI¢tUO .235: I III. . . E III] musixucu: .H H . . A 35:92.8 month: E All. x u I: III. 438.533 5 near»: 2.22.2.3 3.32:... month: so and”? sou... no 3:: :3: 323.3232 2.22.2.3 22:33.. :2. pzuaxo¢.>zu I U $830.35.: pzuxzoc; 47 As the study of image developed the term began to cquire a large number of hyphenated meanings. Positive d negative after—images, memory after—images, recurrent 93,94 ages and tied images were differentiated. Eidetic magery was the center of considerable research.95 Bartlett, in l92l, anticipated the modern use of he term image. He said, ". . . the image . . . appears 3 be most clearly connected with a general affective endency . . . the function of the image . . . is wholly 3 initiate and further some familiar mode of behavior "96 Importance of classical image conCepts'to‘marketing.-— lthough only a tenuous relationship exists between the ecor, "Visual Reading: A Study of Mental Imagery," erican Journal of Psychology, XI (January, 1900), 225; . H. Winch, "The Function of Images," Journal of Philosophy, 3 chology and Scientific Methods, V (1908), 352. 93Bent 8. Russell, "Brain Mechanism and Mental Images," sychological Review, XXVII (1920), 234; G. Dawes Hicks, 0n the Nature of Images," British Journal of Psychology, V (October, 1924), 145; Titchener, op. cit., pp. 73—75. 9“For a list of image terms and definitions used 2 DSYChiatry and psychoanalysis see Holt, op. cit., p. 5. 95Gordon w. Allport, "Eidetic Imagery," British Jour- al of Ps chology, XV (October, 192“), 100; E. R. Jaensch, idetic Imagery New York: Harcourt Brace and World, 1930), einrich Kluver, "An Experimental Study of the Eidetic Ype," Genetic Psychology Monograph, March, 1929. 96F. C. Bartlett, "The Functions of Images," British ournal of Ps chology, XI (April, 1929), 330—331. 48 y r uses of the term image and its modern application Larketing, a number of important aspects should be ad. First, the classical use of image referred to a :eptual phenomenon and the present usage is but one > removed from this. Second, it appears that research effort has not 1 directed towards the sensory aspects of images of iucts, brands, corporations or department stores. A ful basis for this type of study is available in the erature of mental and memory images. Third, the introspective method used in researching tal, memory and eidetic images has at least a familial ationship to some of the projective techniques used in ern marketing research. The contributions of classical ge research to present—day image study are valuable. Image in Modern Social enCe Use of the image concept in modern social science is ilar to the application of this concept in marketing. is employed in a manner which is partially perceptual partially attitudinal. Krech and his associates state, e responses of the individual to persons and things are DGd by the way they look to him——his cognitive world. the image or 'map' of the world of every person is an U9 ndividual one."97 His image is the result of his wants, xperiences and environmental circumstances. Boulding used essentially the same definition of mage in his work, The Image.98 He writes, "The subjective :nowledge structure or image of any individual . . . con— sists not only of images of 'fact' but also images of .n99 ' 'value. According to Boulding a person’s behavior iepends on the image of the world he holds.loo In his iescription of the concept image, Boulding lists various types of images. These are: spatial, temperal, relational, personal, value, affectual and public images.101 Image applications.——The concept of image has been applied to a wide variety of research. It has been used in the study of how various occupations are perceived by 02 college students,1 how the federal government federal 103 employee is perceived. Image has been used to discover 97David Krech, Richard S. Crutchfield and Egerton L. Ballachy, Individual in Society (New York: McGraw—Hill Book Company, Inc., 1962), p. 17. 98Kenneth E. Boulding, The Image (Ann Arbor: The University of Michigan Press, 195 . 99Ibid., p. 11. lOOIbid., p. 6. 101 Ibid., pp. 47-u8. 102D. D. O'Dowd and D. C. Beardslee, College Stu- dents' Images of a Selected Group of Professions and gaggpggiggg (Middletown, Connecticut: Wesleyan Uni— versity, 1960). 103 F. P. Kilpatrick, M. C. Cummings, Jr. and M. K. ;_ 50 04 ‘Navy men perceive the submariner,l and to find how 105 . public views the causes of cancer. In the cor— 'ate setting image has been used to describe how the >1ic views capitalism,106 big business107 and various 1panies, brands and products. Of particular interest ce are the marketing applications of the image concept the analysis of consumer behavior which will be ex— ined in the following chapter. nnings, The Image of the Federal Service (Washington: ookings Institute, 1965), Chapter 10. 1 1 louw. J. E. Crissy and S. Pashalian, "The Inter— .ew," III. Aids to the Interviewer-The Submariner ereot pe (M. R. L. Report, No. 21D, Vol. XI, No. 31, ) ; tober, 1952), ‘ 1 105Hans Toch, Terrance Allen, and William Lazer, Fhe Public Image of Cancer Etiology," Public Opinion Larterly, xxv (1961), pp. 411—414. 106William H. J. Whyte, Is Anybody Listening? Iew York: Simon and Schuster, 1952). 107John W. Riley, Jr. (ed.), The Corporation and is Publics (New York: John Wiley and Sons, Inc., 3 CHAPTER III IMAGERY AND MARKETING The use of the term image in marketing has been imited primarily to a description of a set of attitudes eld by an individual or a group of individuals about a articular product brand, corporation or department store. n addition, the concept of self image has been conceived f as that set of attitudes about self that the consumer olds. Strategically, the notion of image has been used c analyze product offerings in terms of their communality ith the self images of potential consumers. Brand Images A brand image may be thought of as the psychological rapper containing the physical nature of the product. ardner describes it as, "A pattern in the mind which is rejected on the brand . . . a symbol which evokes a com- lex set of meanings and feelings in the mind of the be— older."l’2 This image is the resultant of all the lBurleigh B. Gardner, "Quantitative Research and rand Image," Marketing Concepts in Changing Times, R. M. ill, editor (Chicago: Proceedings of the Conference of he American Marketing Association, December, 1960), p- 55—57. 2See also Donald A. Laird, "How the Consumer 51 r—I—————————_——w—w,..i ,, 52 periences (real, vicarious, or imagined) that an indivi- 11 has had with this and similar brands. Brand images are initially perceptual in nature. a needs, goals and expectations of the individual in- lence his perceptions and so affect the development of 3 image of a particular product. Over time, his experi— ze with the product, his physical and psychological rironment and his social relationships change and mold n3 Ls "dynamic relationship of person to product. Thus, analyzing the nature and importance of brand images, a study must go beyond perception to the individual’s ier drives and his interpersonal conduct. It is the nmation of these factors which results in his tendency respond to particular brands in particular ways. There are advantages for the consumer which accrue m his images of brands. The image a person has of one nd becomes a frame of reference for his analysis of er brands of the same product. Brand images also ilitate the speed and consistency of Judgment in pro— t selection. A number of studies have been carried out on various ets of brand images. Because of the competitive value the results, much of this research is not in the public ain. Wells and his associates, using an adjective imates Quality By Subconscious Sensory Impressions," rnal of Applied Psychology, XVI (1932), 297. 3Gardner, op. cit., p. 55- 53 ack list, described the brand images of 1956 and 1957 :omobiles.Ll In an article, "The Brand Image Myth," ins concluded from his research on, . . . the two best advertised and largest selling automobiles . . . that either images do not exist for these brands or that they are not meaningfully related to purchase behavior. Neither images attributable to the product or images due to personality differences of the owners could be found.5 may be that Evans' use of personality tests to classify a purchasers, rather than classification by self image, 1 to his conclusions. Tyler lists three kinds of brand imagery: sub- :tive, objective and literal. The subjective image :empts to involve the individual with the product by Lf—identification. The objective image tries to sell a product by appealing to emotions. Literal imagery tes the form of visual, e.g., brand name, trademark, kage design.6 A successful brand image would encom- s all three types of imagery. Common perceptual experience with a product among umber of persons results in what Gardner calls the “W. D. Wells, F. J. Andriuli, F. J. G01 and Stuart der, "An Adjective Check List for the Study of Pro— t Personality," Journal of Applied Psychology, XLI 57), 317«319. 5Franklin B. Evans, "The Brand Image Myth," iness Horizons, IV (Fall, 1961), 26 6William D. Tyler, "The Image, The Brand, The Sumer," Journal of Marketing, XXII (October, 1957), 54 >ublic image of a brand. This is "A concensus of mean—w Lngs appearing in the minds of a large number of people . . . exposed to a brand and its communications."7 The :heory of social groups leads us to expect that there vould be not one but a number of public images of a )rand. It is the public image, or images that are of interest to the marketer. He must concern himself with :he elements of the image of his brand that are held in :ommon by a number of potential consumers. It is on this aggregate image that he must base his marketing strategy. Self Image Newman describes the buying process as the, "Match— ing of a person's self image with the image of a product 8 )r brand." This concept of consumer behavior is in 1greement with the discussion of the self concept and notivation in Chapter II. At that point the self was iescribed as both the organizer and the object of wants and goals related to self enhancement. Grubb states that, 'The term 'self concept' denotes the totality of an indi— Iidual's attitudes, feelings, perceptions and evaluations 7Gardner, op. cit., p. 57. 8Joseph W. Newman, Motivation Research and Marketing EEEEEEEEE (Boston: Harvard University, Graduate School 3f Business Administration, Division of Research, 1957), L 55 f himself."9 Self evaluation is essentially a comparison f self with significant others. This is important to the arketer because of the role it plays in purchasing be— avior. Grubb argues that goods are, "Symbols serving as eans of communication between the individual and his ignificant references" (groups and individuals with whom e identifies).lo The marketing executive must present is goods in such a way as to enhance the self esteem of hose who purchase them. Newman reported the results of a dieting study which ndicated that, "It was the woman's evaluation of her eight (in terms of how she felt others saw her weight) ather than her actual weight which largely determined her "11 The Chicago ttitude and behavior toward dieting. ribune study of what automobiles mean to Americans indi— ated that the car is more than a means of transportation. t is also a status symbol and a means of expressing 9Edward L. Grubb, "Consumer Perception of 'Self ancept’ and Its Related Brand Choices of Selected Pro— 1ct Types," Marketing and Economic Develgpment, Peter D. ennett, editor (Chicago: Proceedings of the Conference ‘Zhe American Marketing Association, September, 1965), 19. 10Ibid., p. 420; See also, S. J. Levy, "Symbols Which We Buy," Advancing Marketing Efficiency, Lynn Stockman, editor (Chicago: Semi-Annual Conference of erican Marketing Association, December, 1958), p. 410. llNewman, op. cit., pp. 320—342- 56 dividuality. In other words the car is used as a symbol r enhancement of the person's self image.12 The theory of cognitive dissonance supports the 13 To obtain relief portance of the notion of self image. om post-transaction dissonance an individual may turn to hers with whom he identifies for approval and support r his purchase. According to Festinger, social support duces dissonance.lu’15 The dynamic interrelationship of the image of self, 3 image of a brand and the image of the place of purchase 16 ays an important role in the consumer buying processes. marketing executive must be cognizant of this relation— .p and exploit it in the development of his strategy. Corporate Image The aggregate image of a corporation is not simply iescription of the company, its facilities, products, >p1e and communications. It is rather an interpretation, ‘ sed on the perceptions, motivations and social l2Ibid., pp. 221—261. 13Leon Festinger, A Theory of Cognitive Dissonance anston, Illinois: Row Peterson, 1957). lAIbid., p. 188. 15See also, Bruce C. Straits, "The Pursuit of the sonant Consumer," Journal of Marketing, XXVIII (July, 3), pp. 62—66. 16James A. Bayton, "Motivation, Cognition and Learn- --Basic Factors in Consumer Behavior," Journal of :eting, XXII (January, 1958), 282-289. 57 relationships of individuals, of these facets of the cor— poration. The resulting set of attitudes clustered around the corporation is its image. The aggregate image is the summation of these consumer images. Riley terms the corporate image the organizational counterpart of the self image.17 This statement is significant in two ways. First, it points out two oppos— ing perspectives of the image of a corporation. Second, it points out the tendency to endow the corporation with human qualities. The corporate image may be seen from the perspective of those within the organization. From this vantage it is usually viewed as something created as a matter of policy and action. Martineau points out a problem of this per— spective. In virtually every area we have studied, we find the biggest discrepancies imaginable between what a company thinks of itself——the image it be— lieves it is presenting to the public—~and the way the consumer actually sees it.1 The corporate image can be viewed from another per— Spective——that of the consumer. The image is seen as the property of the individual perceiving the company, rather I 1 17John W. Riley, Jr., "The Nature of the Problem," The Corporation and Its Publics, John W. Riley, Jr., 3ditor (New York: John Wiley and Sons, Inc., 1963). p. 1. 18Pierre Martineau, Motivation in Advertising: M2££1§§_That Make People Buy (New York: McGraw—Hill Book Company, Inc., 1957?, p. 1714’. 58 ;han, as a property of the company. This conception of :he corporate image more nearly agrees with perception ind attitude theory. Interpreting the corporate image as the counter- >art of the self image also points out the human tendency :o anthropomorphize. In replying to questions about com— >anies people tend to attribute to the firms human quali- ;ies such as warmth—coldness, friendliness—unfriendliness, sophistication—lack of sophistication, and imaginative— 1nimaginative. This is reflected in the use of the terms éeputation and personality in reference to companies in 9 . l Lmage research. Social Factors Affecting :he Corporate Image There are a number of social factors important to :he understanding of the corporate image. The dynamics )f the relationship between an individual's perceptions, ittitudes and group memberships hold true for images. dartineau states, . . . the corporation is addressing itself to many different publics, each of which is looking at the corporate image from behind a different 19Pierre Martineau, "The Personality of the Retail Store," Harvard Business Review, XXXVI (January—February, 1958), 52—53, Stuart U. Rich and B. D. Portis, "The Imageries of Department Stores," Journal of Marketing, (XVIII (April, 196A), 15; Leon Arons, "Does TV Viewing Influence Store Image and Shopping Frequency"? Journal 3f Retailing, XXXVIII (Fall, 1961), 11. m 59 set of lenses . . . they see the image differently because their perceptions, their expectations and their wishes differ.20 a similar vein Heidingsfield writes, the, . . . interpretation of an image depends upon the personality, socio-economic status, current emotional well—being and culture of the consumer. Individuals react according to the conditioning of their social group . . . For the purpose of identifying aggregate images the )lic can be divided into groups on the basis of their 2 . . . 2 Lationship to the corporation customers——non—customers, . 2 (a1 customers-—occaSiona1 customers, 3 employees-—non— )loyees. The public can also be divided on the basis membership groups and statistical groups: social asses, occupational groupings, age categories.2u The periences, expectations and values of the members of , 2OPierre Martineau, "Sharper Focus For the Cor— rate Image," Harvard Business Review, XXVI (November— :ember, 1958), 53. 21M. S. Heidingsfield, "Building the Image: An sential Marketing Stratagem," New Directions in Market- 5, F. B. Weber, editor (Chicago: American Marketing sociation Proceedings, June, 1965), p. 135. 22Martineau, op. cit.; L. M. Harris and Mass Ob- rvation, Ltd., Buyers Market: How to Prepare for the N Era in Marketing (London: Business Publications, 3'" 196?), p' 79‘ 23 Loewer, op. cit., p. 196. 2”See for example, Heidingsfield, op. cit., p. 1; Harris, op. cit., p. 85; Rich and PortiS, op. cit., 12; Loewer, op. cit., p. 198. 60 ach of these groups will result in differences in the ggregate images they hold of a particular company. he Corporate Image As A tereotype Tucker researched the question, how much of the orporate image is stereotype? He concluded that be- ause, ”. . . the correlation coefficients (were) high etween bank images and other businesses . . . (that) . . there may be a stereotype involved."25 A year ater, Hill reported on work done by National Analysts rom which he asserted that corporate images, ". . epresent distinct corporate personalities and are only n part due to stereotypes, which are 'halo' images."26’27 he conflict seems to be one of definition and emphasis. he image one holds of a corporation is a stereotype in hat it is an oversimplification of reality. Also, be— ause companies have many common characteristics, many of he elements of the images of these companies will be imilar. However, it is the dissimilarities that are ignificant to the marketer. For it is those facets of 25W. T. Tucker, "How Much of the Corporate Image 8 Stereotype"? Journal of Marketing, XXV (January, 1961), 2. 26Edward W. Hill, "Corporate Images Are Not Stereo— ypes," Journal of Marketing, XXVI (January, 1962), 75. 27Pierre Martineau, "The Personality of the Retail gogi," Harvard Business Review, XXXVI (January—February, 5,53 61 e company that consumers perceive as different that ve it its distinctive image. Department‘Store'Image. The department store image is just a special case the corporate image. The difference lies in the *eater degree of intimacy and awareness that consumers 1V6 of department stores. Consumers relate directly .th department stores, their physical premises, employees 1d products. Thus a person develops more intimate images ? department stores based on more first-hand information 1an he does about other corporations. Harris, commenting on the British retail scene, ascribes the store image in terms of "atmosphere." He ?ites, ". . . the impressions the shop makes upon her :he customer) will be synthesized to form the subjective H28 ality—-atmosphere. Although Harris agrees that many ctors influence the consumer image of a store he states, n the last resort, a 'good' atmosphere depends upon the titude of the staff . . ."29 The negative side of the lationship between staff members attitudes and actions d the consumer's image is commented on by Collazzo: An important source of mistrust arises out of dealings with the person, who is, as far as most customers are concerned, the store's image. That person is the clerk. When he is incompetent, the 28Harris, op. cit., p. 83. 291bid., p. 107. 62 ‘ store is incompetent, when he is inconsiderate the store is inconsiderate, when he high pres— sures the customer, the store is high pressur— ing the customer.30 The consumer's image of a department store is luenced by a multitude of factors. A shopping study iich uncovered reasons women enjoy shopping in de— sment stores. The reasons, in order of their impor— :e (based on the percent of sample reporting reason), (a) recreational and social aspects; (b) seeing new things, getting new ideas; c pleasant store atmosphere, displays, excite— ment; (d) bargain hunting, comparing merchandise, spend— ing money; (e) acquiring new clothes and household items, and 31 (f) helpful salesclerks and other store services. It is interesting to note that only 3 percent of the ple of A,500 women in New York and Cleveland gave help— 1store clerks as a reason for emjoying shopping. How— , 20 percent of the sample indicated that discourteous inefficient salesclerks made them dislike shopping. 3r reasons for disliking shopping in order of impor— 3e are: 30Charles J. Collazo, Jr., Consumer Attitudes and strations in Shopping (New York: National Retail Chants Association, 1963), pp. 113-114. 31Stuart U. Rich, Shopping Behavior of Department re Customers (Boston: Harvard University, Graduate oo1 of Business, Division of Research, 1963), p. 66. ‘ \‘7m)' 63 (a) crowds, boredom, fatigue; (b) inconvenience in making arrangements for getting to store; (c) poor or confusing array of merchandise, difficulty in finding what you want; (d) deciding what to buy, spending money, high prices, and (e) discourteous or inefficient salesclerk service. Another report based on the same study indicated at salesclerk service was of prime importance for con- nience to women shoppers.33 The results of this study dicate that female consumer behavior is very much af- cted by the attitudes and actions of department store lesclerks. These likes and dislikes of department store shopping e similar to Collazzo's satisfactions and frustrations of opping3u and Aron's store qualities.35 Fisk summarizes ese store qualities under the heading cognitive dimensions the department store image36 (see Table l). 32Ibid., p. 66. 33Stuart A. Rich and Bernard Portis, "Clues for . tion, From Shopper Preferences," Harvard Bu81ness ReView, I (May—April, 1963), 1A7. 3”Collazzo, op. cit., pp. 68-69. 35Arons, op. cit., p. 10. 36George Fish, "A Conceptual Model for Studying nsumer Image," Journal of Retailing, XXXVII (Winter, 61-1962), 5. 6A \BLE 1.—-Department store image. Cognitive D . Dimensions eterminants Locational (a) access routes (b) traffic bar— Convenience riers (c) traveling time (d) parking availability Merchandise (a) number of brands stocked (b) Suitability quality of lines (o) breadth of Value for price Sales effort and store services . Congeniality . Post—trans— action satis— faction assortment (d) depth of assortment (e) number of outstanding departments in the store (a) price of a particular item in a particular store (b) price of same item in another store (0) price of same item in substitute store (c) trading stamps and discounts (a) courtesy of sales clerks (b) help— fulness of salesclerks (c) reliability and usefulness of advertising (d) bill- ing procedures (e) adequacy of credit arrangements (f) delivery promptness and care (g) eating facilities (a) store layout (b) store decor (c) merchandise displays (d) class of customers (e) store traffic and congestion (a) satisfaction with goods in use (b) satisfaction with returns and ad— justments (c) satisfaction with price paid (d) satisfaction with accessibility to store 65 Although numerous facets of department store oper— ations may influence a consumer's image, it is not neces- sary for a person to have any actual experience with a >articular department store to have an image of it.37 38 Second hand information or imagination are sufficient. Phe Influence of Experience 3n Department Store Image What affect does experience have on an individual's image of a department store? There are two aspects of :he answer to this question. On the one hand, Collazzo ?eports from his research that, ". . . attitudes become Less favorable as shopping becomes more frequent and routine."39 So, for the individual, repeated experience arings boredom and negative attitudes. On the other hand, {arris quotes a British housewife who says, "You get used :0 one shop and you feel at home there . . . but in a fresh shop you don't know where you are and that takes "40,41 ill the fun out of shopping. These two positions are reconcilable because there are so many other factors which affect the individual's image. It is logical that ‘ 37Harris, op. cit., p. 79; Martineau, "Sharper locus for the . . . , op. cit., p. 53. 38Heidingsfield, "Building the Image . . . ," 3p, cit., p. 138. 39Collazzo, op. cit., p. 104. uoHarris, op. cit., p. 100. ulSee also, Arons, op. cit., p. 9. 66 peated experience will bring about a change of atti— ies because of the increased ability of the person to scriminate among retail institutions. Martineau comments on the images held by customers 1 non—customers. Customers: . like the products, they are familiar with them, they read the advertising to support their favorable opinions. But non— customers very often have negative stereotypes of the company which prevent them from learning anything about the products 42 aerson's behavior as a consumer is affected by his image 1 his behavior tends to lend support to that image. nparison of viewers of Montgomery Ward's advertisements :h non-viewers by Arons, showed, "A general shift in a direction of more favorable image . . . (by viewers) . ."43 Experience with any of the facets of a depart- It store results in some change, positive or negative, the image held by the consumer, and ultimately in his ring behavior. 3 Affect of Needs and Social assures on Department Store HQ: Needs play a role in the creation of an individual's H 1ge of a department store. Collazzo emphasizes, . . 3 abilities and needs of consumers as determined by u2Martineau, op. cit., p- 53- u3Arons, op. cit., p. 11. 67 their stage in the life cycle, economic status and experience, condition their attitudes to retail situ— ations . . ."44 In addition, individual's tend to form an image of a department store in accordance with the conditioning of their group. The norms, values and attitudes of the groups with which a person identifies 45,46 shapes that persons image. Intimate primary groups, secondary groups and even statistical groups have an im- pact on the image a person has of a department store. Collazzo cites Rainwater's study of the working man's wife which indicates that lower class women feel less secure and are more pessimistic about the future. "This affected their attitudes towards charge accounts, sales- ."47,48 From his people, strange stores and brands own research Collazzo reports that family size, stage in the family life cycle and economic and social conditions uuCollazzo, op. cit., p. 63. uSAlfred Oxenfeldt, David Miller, Abraham Shuchman and Charles Winick, Insights Into Pricing . . ., op. cit., p. . u6See also, Levy, op. cit., pp. 413—414. u7Collazzo, op. cit., pp. 13—14. u8For the complete study see Lee Rainwater, Richard Coleman and Gerald Hander, Workingman's Wife (New York: Oceana Publications, 1959)- ,g. 68 act attitudes toward shopping and retail insti— ions}19 The importance of social class in influencing con— er attitudes about department stores is reported by 1s,50 Harris,51 Martineau52 d.53 and Heidingsfiel ;ineau in discussing a number of Chicago Tribune lies of spending—saving behavior, retail store loyal— ; and commodity tastes concluded, and I agree, that, L of these studies reveal close relation between Lee of store, patterns of spending and class member— )."54 Gardner, however, disagrees and states, "The ;e of a . . . (department store) . . . does not vary 1 group to group . . . (i.e., social classes) . . . users to non users."55 The majority of the available ience appears to refute Gardner's position. ugCollazzo, op. cit., pp. 72—73, 103- 50Arons, op. cit., p. 95- 51Harris, op. cit., p. 85- 52Martineau, "The Personality of the Retail Store," cit., p. 50. 53Heidingsfield, op. cit., p. 141. 5”Martineau, "Social Class and Spending Behavior," cit., p. 126. 55Burleigh B. Gardner, "Behavioral Sciences As. ited to Image Building,” New Directions in Marketing, 3. Webster, editor (Chicago: American Marketing DCiation Proceedings, June, 1965), p. 147. 69 mage of Self and Department tore‘Image The image of self plays a role in the development f an individual's image of a department store and his uying behavior. Martineau writes, "The shopper seeks he store whose image is most congruent with the image he has of herself."56 And a good part of the image of store in the Shopper's mind is composed of what kind f people shop there.57 Loewer, in a recent doctoral dissertation at manford University, using Newman's notion of the buying )rocess, studied the relation between an individual's .mage of self and his image of the totality of the de- >artment store. He found that: ". . . at least as far is occupational categories were concerned persons tend to .dentify to a far greater degree with those stores where :hey perceived people like themselves to be customers."58 ma e of Branches of e artment Stores The image of a multi-unit department store is more omplex because of interrelationships with the images of 56Martineau, ”The Personality of the Retail tore," op. cit., p. 57See also Levy, op. cit., p. 410. 58Robert A. Loewer, "A Study of Consumer Per- eption of Department Stores and Department Store Prices," npublished doctoral dissertation, Stanford University, une, 1965, p. 198. 70 branch stores. Heidingsfield warns that, "If there more than one outlet the corporate image must be con— tent with all of its outlets . . . unclear imagery notes a tendency on the part of the consumer to ft easily from one store to another."59 The meaning of this warning for market segmentation not clear. Does it indicate that all the branches of ulti-unit department store should be identical regard— s of their particular trading area? Or does it indi— e that the corporate image should be sufficiently ad in scope so as to encompass the differences in the 'ious branch images? From their study Rich and Portis report that: Martineau's belief that the branch stores take on the personality and characteristics of their downtown units was borne out to some de— gree. . . . However, they also found that the images of branch stores are weaker and that there is considerable similarity among suburban branches of downtown stores . . .50 somewhat surprising finding was the consumers praise the wide variety available at the branch stores. The traditional belief that downtown stores at— tract suburbanites for their wide selection of merchandise may no longer be true today. On the other hand, the downtown stores are still ahead of the suburban stores in terms of overall store reputation and reliability. 1 1' 59Heidingsfield, op. cit., p. 139. 60Rich and Portis, op. cit., p. 15. 61Ibid. 71 Summary The emotional filter through which consumers view spartment stores can be one of the strongest competitive [vantages a store has. Over time, the store's image ends to become confirmed in the customer's eyes because ‘ the selectivity of perception and the relatively fixed iture of attitudes. In fact, the image may act so that 1e store, "Gets credit for all sorts of things which ce contrary to truth."62 An individual's image of a department store also ends to last because it is based on his needs and group ffiliations which are slow to change. The image tends 3 be a stable phenomenon even in the midst of changes in he store itself and in its physical and psychological nvironment.. This stability is an important competitive 001. The department store image, especially of multi— tore, multi—city organizations, tends to give the con- mer something familiar to identify with. In the modern bile society, this is of importance to both department tore executives and consumers. The aggregate department store image is the sum- ation of and so affects those factors which affect the onsumer's image. It is upon this resulting totality f images that department store policy should be eveloped. 62 mage," Op. cit., p. 53. Martineau, ”Sharper Focus for the Corporate CHAPTER IV RESEARCH DESIGN: SAMPLE AND QUESTIONNAIRE This chapter is comprised of four sections. The st covers sample design: sampling considerations, ple type, sample size and sample selection method. second contains a description of the questionnaire: semantic differential technique, selection of de— tment stores, department store shopping questions and ,ographic questions. The third describes the field ’k: selection, training and control of interviewers. . fourth presents the sample responses: factors af—> .ting sample composition and selected demographic Q :racteristics of the sample. Sample Design 1 The sample was designed to test the validity of the othesis that there is a relationship between social J ss membership and individuals' perceptions of parti— ~ 5 ar department stores. It was selected on the basis two considerations: first, the income of the sub— t's housing unit and, second, the geographic location 72 73 he subject's housing unit within the Detroit Standard opolitan Statistical Area (SMSA).1 The sample is a random, stratified, multi-stage sample. The population from which the sample was n is the Detroit SMSA, which includes Macomb, and and Wayne counties and had a population of 1,000 and 1,156,200 occupied dwelling units as of l, 1962.2 A sample size of 720 was determined to re that no subsample would include less than thirty rvations. The sample does not include people in institutions, (itals, nurses' homes, rooming houses, military and er types of barracks, college dormitories, fraternity sorority houses, convents and monasteries, or persons belong to a family of which a member is a department E re employee. ‘ Appendices A, B, C and D contain information on ( sehold size, the presence of children, household in— e and household tenure of the population. ___ 1U S. Bureau of Census, U. S. Census of Population Housing: 1960A Census Tracts, Final Report PHC (1) .shington: U. S. Government Printing Office, 1962. 2"Population and Occupied Dwelling Units in the -roit Region, July 1, 1961," Report of the Population . Housing Committee of the Regional Planning Commission Egoit Metropolitan Area Planning Commission, December, 74 Sample Selection Method The sample selection method was designed to select subjects in five income categories with a degree of geo— raphical dispersion over the Detroit SMSA. Census racts,3 were randomly drawn to conform with nine geo— 5 raphic strataLl and five income strata. Blocks were chosen randomly from the tracts on a systematic basis.6 ousing units7 within the blocks were chosen in clusters on a random basis. The following are the steps in the selection pro— cedure: 1. All of the census tracts in the Detroit SMSA were coded according to income, using the 3 U. S. Bureau of Census, op. cit. “The nine geographic strata were based on the trading areas around the nine J. L. Hudson Company de— partment stores. 5Detroit News Salesman's Map (Detroit, 1961). 6U. S. Bureau of Census, U. S. Census of Housing: 1960, Volume III, City BlocksL Series HC (3), Number 204 (Washington: U. S. Government Printing Office, 1962). 7Ibid., p. 1. A housing unit is defined as "A house, an apartment or other group of rooms, or a single room. . . . When it is occupied or intended for occupancy as separate living quarters, that is, when the occupants do not live or eat with any other persons in the struc— ture and when there is either (1) direct access from the Outside or through a common hall or (2) a kitchen or cooking equipment for the exclusive use of the occu- pants." 75 Detroit News Salesman'S'Map8 as a guide, on a Census Tract Map.9 2. The nine J. L. Hudson department stores were located on the Census Tract Map. 3. Six—mile circles were drawn around each J. L. Hudson store (the six—mile circles were utilized because this is the normal practice of Hudson's management in describing an approximation of each store's trading area). 4. From the trading area for each store two census tracts of each income classification were ran— domly selected. This process resulted in the selection of 90 census tracts, 18 from each in— come classification, geographically distributed around the J. L. Hudson stores in the Detroit ( SMSA. 5. Using census data,10 each of the blocks in the selected tracts was equalized and numbered, omitting blocks with no housing units. This was accomplished by arbitrarily adding together or dividing blocks so that each has approxi— mately 25 housing units. The resulting 8Detroit News Salesman's Map, op. cit. 9U. S. Bureau of Census, Census Tracts, 1960, Op. cit. 10 Ibid., pp. 1—156. 76 operational blocks were then numbered conse- cutively. 6. One operational block was selected from each tract by using a table of random numbers. Thus 90 operational blocks were selected. 7. Using city directoriesll each housing unit in the selected operational blocks was listed and numbered starting at the southwestern corner 12 The rules used of the operational block. in numbering the housing units were: a. Commercial establishments were omitted. b. If more than one city block was included in the operational block, the housing units in each city block were numbered from their llPolk's Detroit East Side Directory 1963 (Detroit: R. L. Polk and Company, 1963); Polk's Detroit West Side Directory 1964 (Detroit: R. L. Polk and Company, 1964); Polk's Birmingham Directory 1964 (Detroit: R. L. Polk and Company, 1964); Polk's Lincoln Park Directory 1964 (Detroit: R. L. Polk and Company, 1964); Polk's Pontiac Directory 1962 (Detroit: R. L. Polk and Company, 19625; Polk's Dearborn Directory 1965 (Detroit: R. L. Polk and Company, 1965); Polk's East Detroit Directory 1958 (Detroit: R. L. Polk and Company, 1958); Polk's Down River Directory 1965 (Detroit: R. L. Polk and Company, 1965); PoIk's Ferndale Directory 1959 (Detroit: R. L. Polk and Company, 1959); Polk's Wayne Inkster and Garden City Directory 1960 (Detroit: R. L. Polk and Company, 1960); Polk's Royal Oak Directory 1960 (Detroit: R. L. Polk and Company, 1960). l2Leslie Kish, Survey Samplin (New York: John Wiley and Sons, Inc., 19 5 , p. 352. Kish estimates that, "Directories omit 5 per cent or less on the average” of housing units. 77 respective southwestern corners in a consecutive list following from the lowest to the highest block number as designated by the Census.13 0. If a city block contained more than one operational block, the housing units were numbered in groups of 25 beginning at the southwestern corner of the city block. d. In buildings containing multiple housing units the units were numbered from the lowest number or letter to the highest as if they were separate numerical addresses. e. In cases where there was no complete list- ing of the housing units for a selected city block a field enumeration was carried out. 8. From the housing unit list for each operational block eight dwelling units were selected at random with the use of a table of random numbers. 9. This process yielded 720 subject addresses, 144 from each income classification, geographically distributed around the Hudson stores in the Detroit area. 13U. S. Bureau of Census, U. S. Census of Housing ‘ I a, QB. Cit. 78 10. It was estimated that approximately 30 percent of the selected subjects would not respond because of refusals to cooperate and "not at homes." To allow for this each interviewer was given a systematic method for randomly selecting additional subjects in his oper— ational blocks so that the quota of eight sub— jects per selected operational block would be accomplished. 11. The subject within the housing unit was the male or female head of the household. It was assumed that because of availability, the majority of the subjects would be female. The interviewers were instructed to obtain one male subject for every three female subjects. It was anticipated that this process would result in 1 sample with approximately equal numbers of subjects in each social class category.114 1“Subjects were assigned to social classes by neans of Warner's Index of Social Characteristics adminis— tered as part of the interview. The social class group— ings used were adapted from Warner's six strata. They were: upper, upper middle, lower middle, and lower. W. Lloyd Warner, Social Class in America (New York: Harper and Row Publishers, 1960), Chapters 8 and 9. 79 15 The Questionnaire The questionnaire is made up of three parts each with a distinctive purpose. The first part (questions 1-11) contains a number of semantic differential scales designed to measure the subjects' attitudes towards three actual department stores, one hypothetical "ideal" department store, and the activity of shopping. The second part (questions 12—17) is composed of a set of questions to determine the subjects' actual department store shOpping behavior. The third part (questions 18- 27) is a series of questions designed to obtain demo— graphic information about the subjects. The Semantic Differential The semantic differential was developed by Osgood and his associates16 as a tool for measuring meaning. The purpose of our factor analytic work is to devise a scaling instrument which gives repre- ‘ sentation to the major dimensions along which meaningful reactions or judgments occur.17 A significant by—product of Osgood's work in experi— mental semantics was a new rationale and approach to the measurement of attitudes. In his research on meaning, 15See Appendix E. 16Charles E. Osgood, George J: Suci and Percey H. Tannenbaum, The Measurement of Meaning (Urbana: Uni- versity of Illinois Press, 1957 : l7Charles Osgood and George J. Suci, "Factor Analysis of Meaning," Journal of Experimental Psychology, 40 (May, 1955) , 325 . “Mm—V 80 :titude was identified, "As one of the major dimensions f meaning in general ."18 Osgood and his colleagues sed the term attitude to mean a: ”Learned implicit pro- ess which is potentially bipolar, varies in its intensity, ad mediates evaluative behavior . . ."19 Attitude then, 3 part of the internal mediational activity that oper— tes between most stimulus and response patterns. Attitude is identified with the evaluative dimension f meaning. In terms of the operations of measurement with the semantic differential, we have defined the meaning of a concept as its allocation to a point in the multi dimensional semantic space. We then define attitude toward a concept as the projection of that point onto the evaluative dimension of that space.20 Measurement of attitudes towards concepts with the emantic differential is accomplished by having subjects ake repeated judgments of that concept against a series f descriptive bipolar scales. These judgments are made n a seven point interval ordinal scale. The intervals on the scales in the questionnaire ere labeled to assist the subjects in rating the concepts. CONCEPT A ”‘1‘“! A Cold dly Store : : —-—-—-———' : -—-—-"' : ——'—‘ ’ ' ————-—— Unfriendly Store Very Quite Slightly Neutral Sligr.tly Qui’c ler-y l8Osgood, The Measurement of Meaning, op. cit., p. 89. “1.1.11...” ~— 20 191bid., pp. 189—190. Ibid., p. 190. 81 By assigning numerical values to each interval on 3e scale the qualitative data may be quantified. In nis way both the direction and intensity of an atti- 1de toward a concept may be measured. For tabulation purposes the intervals were numbered rom one through seven from what was judged to be the ast desirable descriptive phrase (from the point of iew of department store management) to the less prefer— ble descriptive phrase. dvantages and Limitations of he Semantic Differential As was stated in Chapter II, the semantic differential as high test-retest reliability and high validity. The ollowing are some of the advantages of using the semantic ifferential for marketing research: 1. It is a quick efficient means of getting, in a quantifiable form and for large samples, the direction and intensity of attitudes toward a concept. 2. It provides a comprehensive picture of the image. 3. It represents a standardized technique for getting at the multitude of factors which go to make up an image. 4. It is easily repeatable and quite reliable. 5. It avoids stereotyped responses and allows for individual frames of reference. 82 6. It eliminates some of the problems of question phrasing, such as ambiguity and some forms of question bias. It also facilitates the inter- viewing of respondents who may not be too articulate in describing their reactions to such an abstract concept as their image of a corporation.21 There are a number of limitations in using the semantic Lfferential in marketing research. 1. It is an attitude scale and so does not allow direct prediction of behavior.22 2. It does not measure much of the content of an attitude in the denotative sense.23 3. It is difficult to select suitable scales for the concepts in question. 4. Gatty and Allais caution that: ”There remain some questions of fundamental validity ."ZM’ES 21William A. Mindak, "Fitting the Semantic Differ- itial to Marketing Problems," Journal of Marketing, 25 April, 1961), 28—29. 22Osgood, The Measurement of Meaning, op. cit., D. 198—199. 23Ibid., p. 195. 2“Ronald Gatty and Claude Allais, The Semantic ifferential Applied to Image Research (New Brunswick, aw Jersey: Department of Agricultural Economics, ltgers University), pp. 3—4. 25See also, Gerald Zaltman, Marketing: Contributions ?Om the Behavioral Sciences (New York: Harcourt, Brace ad World, 1965), pp. 108-109. 83 A number of modifications which might be incor— >rated to make the semantic differential more useful in ereting research were suggested by Mindak. Descriptive >uns and phrases (in addition to Osgood's adjectives) aveloped from pretesting for the study of a particular )ncept may make the results more meaningful. Phrases ty be used which, although not denotatively opposite, rem to fit logically in people's frame of reference. 1e use of these connotative or non—polar opposites may .iminate clustering about the middle of the scales which asults from respondents' hesitancy to use the negative ;de of the scales. Built—in control concepts, such as 1e "ideal" company or product may be used to test con- 26 apts or competitive concepts. alection of Descriptive :ales The selection of the bipolar descriptive scales for 18 questionnaire resulted from an analysis of studies of apartment stores reported in the literature. These :udies were concerned with the satisfactions and frus- ‘ cations of consumers,27 shoppers likes and dislikes of 26Mindak, op. cit., pp. 29—30. 27Collazzo, op. cit., Chapters 6—10. 84 epartment stores,28 and the images and personalities of epartment stores.29 Individual scales were designed to gather data bout four department store image dimensions: congeniality f the store, sales personnel, locational convenience, and erchandise characteristics.3O Thirty bipolar descriptive cales were selected to measure the images of the depart— ent stores. election of Department tores The department stores selected for the study were hosen on two considerations, geographical location and ocio—economic position. It was desired that each depart— ent store have a number of branch stores throughout the etroit SMSA. This would allow subjects in all geo- raphic locations an approximately equal opportunity to 31 e acquainted with each department store. The stores 28Rich, Shopping Behavior of Department Store ustomers, op. cit., pp. 65—65: 29Martineau, "The Personality of the Retail Store," p. cit., pp. 51—52; Stuart U. Rich and B. D. Porter, The Imageries of Department Stores," Journal of Market- 95 (April, 1964), 10—15; Grey Matter (Retail-Edition)— September, 1965), 11, no. 5 (New York: Grey Advertising, ncorporated, 1965); "The Detroit Consumers' Beliefs and onceptions Associated with the J. L. Hudson Company,” eported in a J. L. Hudson Company memo, March 30, 1959. . 3OThese image dimensions were adapted from George lsk, "A Conceptual Model for Studying Customer Image," OUrnal of Retailing, 37, N0. 4, p. M 31 The J. L. Hudson Company had nine stores located in 85 >sen were: the J. L. Hudson Company, Sears Roebuck and lpany and Federal Department Store, Incorporated. Stores :e selected to represent high, medium and low on a socio- >nomic continuum. It was judged that Hudsons would tend place high on this continuum, Federals low and Sears in 9 middle. A hypothetical "ideal" department store was also ad in the study as a control concept.32 Ratings on : "ideal" department store were obtained in order to :e comparisons with the test stores. To eliminate bias caused by order, the pages of the astionnaires were arranged so that each store appeared “st, second, third and fourth in an equal number of astionnaires. : Activity of Shopping Five bipolar descriptive scales were selected to lsure the consumers' attitudes towards the activity of >pping. The subjects were asked to disregard grocery )pping in rating this concept. The scales were based 33 reports of shopping studies. This information was Ls area; Federals had twenty—five stores and Sears abuck had twelve stores. 32Mindak, op. cit., p. 30. 33Collozzo, op. cit., Chapters 8 and 10; Rich, Jpping Behavior of Department Store Customers, op. cit., ipter 6. .4‘j 86 >btained in order to compare the images of the department stores held by those who enjoy shopping with those who do not enjoy shopping. Shopping Behavior Questions Questions 12 through 17 were designed to obtain in— ?ormation with regard to the respondents' department :tore shopping behavior, especially at the J. L. Hudson stores. In question 12 the subject was requested to list 11s or her first, second and third choices in department stores. This information was obtained in order to com- )are the images of each department store held by those vho pick that store as their favorite as compared to the Lmage of that store held by consumers who pick other stores as their favorite. It was expected that the use of :hree test stores in the semantic differential section of :he questionnaire would create some bias in the listing of ?avorite department stores. Question 13.1 was devised to find which J. L. Hudson store the respondent shops at most often. (Questions 13, L4 and 15 were omitted if the subject claimed she had lever shopped at Hudsons.) This information was obtained in order to compare the image of Hudsons held by people who shop most often at one Hudson outlet with the image 3f Hudsons held by people who shop at another Hudson outlet. 87 Questions 13.2, 13.3, 13.4 and 14.1 were designed acquire information regarding the loyalty of the re— )ondent to the J. L. Hudson Company. These data were )tained to compare the image of Hudsons held by loyal loppers with the image held by occasional shoppers. lso, from Question 13.2, the affect of recency of hopping on image may be measured. Question 15.1 was prepared to get information about he respondent's payment practices when shopping at Hud— ons. The purpose of this question was to obtain infor— ation to allow comparison of the image held by cash ustomers and credit customers. Question 15.3 was devised to gather information re— ;arding the location of the respondent's shopping activi- Lies within the J. L. Hudson stores-—upstairs and for lownstairs. The question was intended to allow comparison )f the image held by people who shop only upstairs with :he image held by those who shop only downstairs and with :he image held by those who shop both upstairs and down— stairs. Question 16.1 relates to the respondent's frequency >f reading Hudson's newspaper advertisements. This infor- lation was obtained to compare the image of Hudson's held >y frequent, occasional and non—readers of Hudson's adver- :isements. Question 17.1 was designed to have the subject com— )are his or her feelings about Hudson's with those of his v/ 88 :r friends. The objective of this question was to v a comparison between the images of those who had 11 support and those who did not. Demographic QuestiOns The major purpose of the demographic questions is lassify the respondents into social classes and into 1y life cycle groups. Questions 18, 19 and 20 were devised to obtain in- Lation regarding the respondent's age, marital status the presence of children. This information was ob— 1ed to compare the images of the department stores 1 by people in various stages of the family life Le. The following are the stages used: 1. Young,a singleb 2. Young, married, no children 3. Young, married, with children0 4. Older,d married, with children 5. Older, married, no children 6. Older, sing1e3Ll By young is meant, head of spending unit (male or ale) under 40; (b) by single is meant, head of spend— unit is not married, widowed, divorced, or separated; 34Adapted from J. B. Lansing and J. M. Morgan, asumer Finances Over the Life Cycle," in Lincoln ’k (ed.), Consumer Behavior, Vol. II (New York: New : University Press, 1955), p- 37' 89 ) by children is meant children living at home or away school and supported by the head of spending unit, 1d (d) by older is meant, head of spending unit (male 'female) is 40 years or older. Questions 21.1 through 26.3 were designed to obtain .formation to classify respondents into social class >oups. Warner's Index of Social Characteristics, was led to place respondents into social class groups.35 ie four social classes used were adapted from Warner's .x strata. The four used are: upper (which includes .rner's upper upper and lower upper), upper middle, lower .ddle and lower (which includes Warner's upper lower and »wer lower). L There are three reasons for using four classes ther than six. One, to define a broad enough classifi— tion of the upper group to insure obtaining sufficient bjects. Two, to increase the relative size of the iddle" classes and decrease the size of the "lower" asses. Three, to simplify the analysis and reporting the results. 3 Index of Social -—_ aracteristics The Index of Social Characteristics measures the :io-economic levels of the community. It is based 35Warner, Social Class in America . . ., op. cit., Lpters 8-11. 11.1 . relsmia 90 the prepositions that economic and other prestige :tors are closely correlated with social class and that ese factors must be translated into behavior acceptable members of a given social level if their potentialities ’ rank are to be realized. Warner's methods of measuring social status have - gone without criticism. Such eminent sociologists C. Wright Mills, Ely Chinoy and Ruth Kornhauser have luated his work. He was variously attacked for neglecting the dynamics of stratification, for a value orien— tation which favored the status quo, for the sub— jectiveness inherent in a ranking technique which reflects private opinion of informants and for his failure to clearly articulate his research pro— cedure.3 e of the criticisms invalidate the use of the I. S. C. an indicator of status in this study. The four status characteristics used in the Index 37 They are: e developed in the Yankee City research. lpation, source of income, house type and dwelling a. Each of the status characteristics is measured 1 scale from one (highest) to seven (lowest). The llting scores are then weighted and totaled to give Lndex of social status. 36Harold M. Hodges, Jr., Social Stratification: :s in America (Cambridge, Massachusetts: Schenkman .ishing Company, Inc., 1964), p. 64. 37 W. Lloyd Warner and Paul S. Hunt, The Social of An American Communipy, Vol. I, Yankee City fig (New Haven: Yale University Press, 19415. 91 Table 2 shows the seven point scales for each of e status characteristics. The weighting used to obtain the Index is as llows:38 Occupation 4 Source of Income 3 Housing Type 3 Dwelling Area 2 Table 3, shown on page 93, shows the weighted total tings and the social class equivalents. Questions 21.1, 21.2, 21.3 and 21.4 were designed obtain information to classify the respondents accord— g to the occupation scale in Table 2. ‘ Question 23.1 was devised to obtain data to classify 8 subjects according to the source of income scale in ble 2. Question 25.1 was designed to allow classification the respondents' home on a revised house type scale. a house type scale in the Ipdgg has been criticized for 5 lack of objectivity because the interviewer classified 9 subject's home by subjective judgment.39 To avoid .8 problem the respondents' homes were classified on : basis of their market value (as seen by the respondent), 38Warner, Social Class in America, op, cit., p. 123. 39Hodges, op. cit., p. 99- 92 TABLE 2.-—Sca1es for making primary ratings of the status characteristics of the I. S. Occupation Professionals and proprietors of large businesses Semi—professionals and smaller officials of large businesses Clerks and kindred workers Skilled workers Proprietors of small businesses Semi—skilled workers Unskilled workers map ‘10th Source of Income Inherited wealth Earned wealth Profits and fees Salary Wages Private relief Public relief and non—respectable income \Imwcme House Type Excellent houses Very good houses Good houses Average houses Fair houses Poor houses Very poor houses 0 \IOchcomy—I Dwelling Area Very high; North Shore, etc. High; the better suburbs and apartment house areas Above average; areas all residential, larger than average space around houses; apartment areas in good condition 4. Average; residential neighborhoods, no deteriora— tion in area 3. Below average; area not quite holding its own, _ beginning to deteriorate, business entering 3. Low; considerably deterioriated, run—down and semi—slum ' ’. Very low; slum 1 Warner, Social Class in America, p. 123. 93 TABLE 3.——Social-class equivalents for I. S. C. ratings.l Weighted Total Social Class of Ratings Equivalents l2 — 17 Upper class 18 — 22 Upper class probably, with some possibility of upper—middle class 23 — 24 Intermediate: either upper or upper—middle class 25 - 33 Upper-middle class 34 - 37 Intermediate: either upper— middle or lower—middle class 38 — 50 Lower—middle class 51 - 53 Intermediate: either lower- middle or upper—lower class 54 - 62 Upper—lower class 63 - 66 Intermediate: either upper- 1ower class or lower-lower class 67 - 69 Lower-lower class probably, with some possibility of upper—lower class 70 — 84 Lower—lower class lWarner, Social Class in America, op. cit., p. 127. 94 7 the rent paid. This scheme also avoided the problems ? judging apartment houses. The following is the revised house type scale: Rating Market Value Rent Paid Monthly 1 more than $50,000 more than $500 2 $35,000 — 49,999 $250 — 499 3 $20,000 - 34,999 $150 - 249 4 $15,000 _ 19,999 $100 — 149 5 $12,500 - 14,999 $ 75 - 99 6 $10,000 — 12,499 $ 50 — 74 7 less than $10,000 less than $50 Respondents' addresses were obtained in question 5.1 in order to place each subject's residence in a Nelling area category according to the scale in Table 2. 3 accomplish this the Detroit News Salesman's Map,“O he Social Rating of Communities in the Detroit Area“1 1d personal observation were used. The subjects were classified by social class in rder to compare the images of the department stores held y members of one class with those held by members of ther classes. Question 27.1 was designed to allow a comparison f the images of department stores held by men and women. 40Detroit News Salesman‘s Mgp, op. cit. 41 >cial Rating of Communities in the Detroit Area (1965). United Community Services of Metropolitan Detroit, .A- 95 Question 27.2 makes possible a comparison of the lages of department stores held by white and non—white aspondents. Field Work The questionnaire was administered by means of arsonal interviews at the homes of the subjects. Per— >nal interviews were used because it was believed that 1e length and complexity of the questionnaire required 1terviewer supervision. In addition, a mail survey auld allow the respondent too much time to deliberate 1 the ratings“2 and returns would likely be low. A team of professional interviewers, from a etroit—based marketing research firm, and a group of :udents from a university in Detroit were contracted to > the field work. They were trained to administer the Lestionnaire and given instructions as to their activi- .es while interviewing}3 Each interviewer was equipped with a kit containing: 1. A set of interviewer instructions; 2. Questionnaires; 3. An identification tag; 4. Street maps of Detroit and suburbs with the general location of the subject addresses to be contacted marked; “2Mindak, op. cit., p. 30. u3See Appendix F. 96 5. Subject address lists with a sketch of the city block or blocks within which the subjects were located. The interviewers were instructed to follow address lists in order and to replace subjects only after refusals or after three unsuccessful attempts to contact the sub- ject. Subjects were to be replaced by selecting the second address on the same side of the street in the di— rection of the listing of the addresses. If the newly selected address was already on the address list the interviewer was instructed to repeat the replacement pro- cess. Interviewing was carried out during the months of November and December, 1966, January, February and March, 1967. The interviews were authenticated by contacting 10 percent of the total sample households by telephone. The interviewers were compensated from a research grant made by the J. L. Hudson Company to Michigan State Uni- versity. Sample Responses The sample was not designed to be representative )f the Detroit SMSA. Rather it was intended to in— :lude a total of 720 subjects in five income categories ?epresenting four social class categories. It was anti— Eipated that the sample would be comprised of approxi— lately equal numbers of subjects in the four social 97 classes. The field work resulted in 652 completed usable questionnaires)“l Of those completed only 27 were deemed unusable. Factors Affecting Sample Composition A number of factors had some affect on the composition of the sample. Racial tension, especially in the lower socio—economic neighborhoods, eliminated some of the origi— nally selected interview areas because: (1) many prospeC— tive respondents would not answer the door or refused to be interviewed, and (2) it was difficult to recruit inter— viewers to work in these areas. The Christmas season and inclement weather after the New Year posed problems for the interviewers in finding subjects at home. This problem was amplified by the fact that no interviewing was done on Saturdays, Sundays and evenings. The length of the inter— view (about thirty—five minutes) and the small size of the print used for the semantic differential scales resulted in a number of refusals and interviews terminated before completion. These factors seem to be the most significant in resulting in 395 not at homes (after two call backs) and 254 refusals. “A questionnaire was deemed usable if the re— Spondent completed at least the semantic differential scales for the J. L. Hudson and the "ideal" department stores and sufficient other questions to allow placing the respondent into the various demographic categories. 98 lected Demographic Charac— ristics of the Sample DemOgraphic characteristics of the sample presented e: social class, family life Cycle, race, sex, age, come and education. Table 4 presents the results of classifying the spondents into social class groups by means of Warner's dex of Social Characteristics.“5 The objective of ur social class groups of equal size, unfortunately was t attained. In the very high class suburban areas the mber of refusals and not at homes was very high as ght be expected. Household servants tended to dis- urage interviewers from contacting respondents. In dition, the rigor of Warner's Index, in which only the p 3 percent of the population are classified as upper ass,”6 tended to make it difficult to place people in is class. Consequently, only 13.3 percent of the sample Sponses were classified as upper as compared to the ob— ctive of 25.0 percent. The upper middle and lower ddle classes were represented approximately in accord h the objective of social class groups of equal size. lower class was over-represented with 33 percent of sample being placed in this category. Again the or of Warner's Index, with its strong emphasis on 45Warner, op. cit., Chapters VIII—X. M6Charles B. McCann, Women and Department Store ertising (Chicago: Social Research, Inc., 1957), p. 99 TABLE 4.--Social class of respondents of usable sample responses. a b Social Number Proportion Class 1966—67 1966-67 Upper 87 13.3% Upper Middle 154 23.6 Lower Middle 195 29.9 Lower 216 33.0 Total 652 100% aRespondents were classified into Social Classes by means of Warner's Index of Social Characteristics, see W. Lloyd Warner, Social Class in America: The Evaluation of Status (New York: Harper and Row Publishers, 1960) especially Chapters 8—10. bPercentages may not equal 100 due to rounding. occupation may have resulted in the high proportion of lower class classifications. The classification of respondents' households into stages of the family life cycle is shown in Table 5. It was anticipated, for purposes of analysis, that each of the life cycle stages would include at least 30 sub— jects. The sample includes less than this number in the young single and young married no children stages and so they were combined into one stage. Each of the other family stages contain more than 30 subjects. 100 TABLE 5.——Family life cycle stage of usable sample responses. Family Life Cycle Stage Number Proportion 1. Young, Single 23 3.5% Young, Married, no children 13 36 1.9 5.4% 2. Young, Married, children 214 32.8 3. Older, Married, children 210 32.2 4. Older, Married, no children 127 19.4 5. Older, Single 66 9.9 Total 652 100%8L aPercentages may not equal 100 due to rounding. Table 6 describes the racial characteristics of the sample. The sample is comprised of 87.5 percent white and 12.5 percent non-white subjects. The proportion of white and non—white subjects may have been affected by the racial tension described above. It may also have been influenced by the geographic stratification in the sample selection process which tended to under—represent the "inner city" where most of the Negro population is concentrated. A- 101 DABLE 6.5-Race of usable sample responses. Race Number Proportion White 571 87.5% Non—White 81 12.5 Total 652 100% The data about the sex of the respondents in the sample is presented in Table 7. Of the usable sample responses, 15.5 percent were male and 84.5 percent were female. TABLE 7.——Sex of respondents of usable sample responses. Sex of Respondent Number Proportiona Male 101 15.5% Female 551 84.5 Total 652 100% aNo attempt was made to select respondents by sex Ln such a way as to be proportional to the male—female aatio of the population of the Detroit Standard Metro- )olitan Statistical Area. In Table 8 the age of the respondents from the lsable sample responses is outlined. It-shows that learly 15 percent of the sample is less than 30 years; 102 TABLE 8.--Age of respondents of usable sample responses. Age Number Proportionb Less than 20 years 9 1.4% 20 but less than 30 years 88 13.4 30 but less than 40 years 156 23.8 40 but less than 50 years 208 32.1 50 but less than 65 years 131 20.3 65 years and more 57 8.7 Total 649a 100% aThree respondents refused to answer the age question. bPercentages may not equal 100 due to rounding. almost 56 percent of the sample is between 30 and 49 years; and about 29 percent is 50 years or older. Table 9 describes the education of the head of the household of the usable sample responses. This table indicates that 31.5 percent of the respondents did not graduate from high school; 21.1 percent attended college; and 24.6 percent hold bachelors or graduate degrees. The prOportion in each education bracket was influenced by the sample selection method. Table 10 presents the household income of the usable sample responses. A breakdown of the sample by household income shows 11.9 percent have incomes of less than ....._.A- 103 TABLE 9.—-Education of head of household of usable sample responses. Education Number Proportionb Elementary 71 11.5% Some high school 123 20.0 Graduated high school 130 21.1 Some college 140 22.6 Graduated college 81 13.1 Graduate school 71 11.5 Total 616a 100% a36 respondents refused to answer education question. bPercentages may not equal 100 due to rounding. TABLE 10.-—Household income of usable sample responses. Income Number Proportionb Less than $4,000 69 11.9% $4,000 - 7,999 132 22.8 $8,000 _ 9,999 108 18,5 $10,000 — 14,999 146 25.3 $15,000 — 24,999 83 14.3 More than $25,000 40 6.8 Total 578a 100% a . . 74 respondents did not answer income question. bPercentages may not equal 100 due to rounding, 104 $4,000; 22.8 percent have incomes of $4,000 to $7.999; 18.6 percent have incomes of $8,000 to $9,999; 25.3 per— cent have incomes of $10,000 to $14,999; and 21.1 per— cent have incomes of $15,000 or more. The proportions were strongly influenced by the sample selection pro- cess. Interviewing difficulties and under-representation of the low income "inner city" may have contributed to the small proportion of the lowest income group. Two limitations result in the sample of respondents. First, insufficient subjects in the upper class make com- parisons of the family life cycle stages within this class impossible. Second, the small number of subjects in the young, single and young, married no children stages of the life cycle necessitate the combination of these stages into one. The division of the sample, on the basis of the other demographic characteristics, re— sults in groups sufficiently large for purposes of analysis. CHAPTER V DEPARTMENT STORE IMAGE: AGGREGATE IMAGES OF TEST AND "IDEAL" STORES This chapter is divided into three sections. The first section describes the method of analysis. The second presents some factors which may have affected the aggregate images of the test department stores. The third is a comparison among the aggregate images of the test and "ideal" department stores as shown by the re- sults of the sample survey. Method of Analysis As was stated in Chapter IV the stores were judged on a seven step interval scale against thirty bi-polar semantic differential scales. To quantify the results the steps were numbered from one through seven; one repre— senting the most favorable judgment from a management point of view, and seven representing the most unfavor— able judgment. For purposes of analysis the thirty bi—polar semantic scales were classified by four image dimensions, The four dimensions are: merchandise suitability, sales 105 106 personnel, store congeniality and locational convenience. Assignment of the scales to the dimensions is shown in Appendix G. Each respondent's score for the scales making up each of the four dimensions was totaled and the mean ob- tained. The means were then subjected to two nonpara— metric statistical tests, the Wilcoxon Matched—Pairs Signed—Ranks Test and the Mann—Whitney U Test. The tests are described in Appendix H. The Wilcoxon test was used to measure the differ- ences in scores between test and "ideal" stores for social classes, family life cycle stages and social classes at various family life cycle stages. For example, the Wilcoxon was used to compare the aggregate image of Sears with that of Federal for young married couples with children in the lower middle class. The Mann—Whitney test was used to measure the differences in scores be— tween social classes, between family life cycle stages and between the same family life cycle stage in different social classes. For instance, the Mann-Whitney was used to compare the aggregate department store image of Hudson's held by members of the upper class with that held by the lower class. The Wilcoxon test results in T scores, which are converted to z scores if the number of pairs in a com— parison (omitting ties) is larger than 25. The statistic 107 T equals the smaller sum of like signed ranks. The Mann— Whitney test results in U scores, which are converted to z scores if the number of cases in the larger group is greater than 20. The value of U is given by the number of times that a score in one group in a comparison pre— cedes a score in the other group. Equivalent two—tailed probabilities may be found by referring T, U and z scores to appropriate tables. The null hypothesis in each comparison is that any difference between the means is due to chance. For example, a two— tailed probability value of .025 would indicate that there are only 25 chances in a thousand that a difference as large as that found occurred by chance. The inference then would be that the statistics were not drawn from the same population. The null hypothesis was rejected only if the two-tailed probability was equal to or less than .05, that is, there was at least a 95 percent change of an actual difference. Possible Factors Affecting Aggregate Department Store Images Two significant events occurred during the period of the study which may have affected the image of the J. L. Hudson Company. One, the Hudson's Thanksgiving Day Parade, may have had a positive influence. The other, a shooting of a clerk in Hudson's downtown store, may have had a negative influence. 108 Aggregate Images'of'Test‘and'"Ideal" Department Stores This section presents a comparison among the aggre- gate images of the test and "ideal" department stores by consumers in each social class, in each family life cycle stage and in each family life cycle stage in each social class. Comparison of the Dimensions of Aggregate Department Store Images by Social Glass Table 11 presents a summary of the Wilcoxon T and probabilities associated with the null hypotheses that there are no perceived differences on the image dimensions among the test and "ideal" stores for subjects classified by social class. In general the results indicate that members of each social class have different perceptions of Hudson's, Sears, Federal's and the hypothetical "ideal" department store. Of the 96 between store comparisons 89 yield differences which are significant at the 5 percent level. In all but one comparison Hudson's is rated higher than Sears and higher than Federal's. Sears is rated higher than Federal's in all cases and "ideal" is rated higher than Hudson's, Sears and Federal's in all compari— sons. These data support the assumption, outlined in Chapter IV, that the three test stores could be placed on a continuum with Hudson's ranked highest and Federal's 1(39 TABLE ll.--Summary of Wilcoxon T and probabilities associated with the null hypotheses by social classes for image dimensions between test and ”ideal" department stores. a Social Classes Department Stores Upper Lower Upper Middle Middle Lower Merchandise Suitability Hudsons- z; -6.17 —7.53 —6.90 —4.46 Sears p <.01 <.01 <.01 <.01 Hudsons- z -6.70 —9.86 -lO.77 —9.86 Federals p <.01 <.01 <.01 <.01 "Ideal"- z -5.58 —9.48 —8.67 —7.83 Hudsons p <.01 <.01 <.01 <.01 Sears- 2 -5.34 —7.14 —8.30 -8.21 Federals p <.01 <.01 <.01 <.01 "Ideal"— z —7.66 -1o.26 —lo.67 -lo 18 L Sears p <.01 <.01 <.01 <.01 "Ideal"- z —6.95 -10.53 -ll.54 -12.06 Federals p <.01 <.01 <.01 <.01 Sales Personnel c b Hudsons- 2d —0. 56 36 -l.25 -3.38 p 5754 0.7188 .2112 <.01 Hudsons- z —2.48 -2.45 -5.29 —5.92 Federals p .0132 .0142 <.01 <.01 "Ideal"- z -5.53 -9.19 -7.47 —6.85 Hudsons p <.01 <.01 <.01 <.01 Sears. 2 -3.39 —2.98 —4.96 ~3.43 Federals p <.01 <.01 ’ <.01 <.01 "Idea1"- z —5.99 —9.04 -8.07 —8.30 Sears p <.01 <.01 <.01 <.01 "Ideal"- z —6.41 —9.65 -10.12 —9.74 Federals p <.01 <.01 <.01 <.01 Store Congeniality Hudsons- z: -4.81 —6.57 —3.93 —1.86 p <.01 <.01 <.01 .0628 Hudsons- z -5.86 -8.38 —7.17 —4.57 Federals p <.01 <.01 <.01 <.01 "Ideal"- z —4.20 —7 16 . —5.62 —8.27 Hudsons p <.01 <.01 <.01 <.01 Sears- z -4.90 —4.46 -3.46 —3.91 Federals p <.01 <.01 <.01 <.01 "Ideal"- z -6.76 -9.28 ‘ -7 70 —8 82 Sears p <.01 <.01 <.01 <.01 "Ideal"- z -6.55 -9.39 -9.02 —10.74 Federal: p <.01 <.01 <.01 <.01 Locational Convenience Hudsons- z: -5.23 -6.26 —3.41 —1.33 s p <.01 <.01 <.01 .1836 Hudsons- z -5.69 —6.58 -3.55 —3-23 Federals p <.01 <.01 <.01 <.01 "Ideal"- z -6.48 -9.07 —9.05 -10.10 Hudsons p <.01 <.01 <.01 <.01 Sears- 2 —3.43 —0.29 —O.74 -2.7l Federals p <.01 .7718 .4592 <.01 "Ideal"- z -7.32 -10.22 , —lo.26 —10.34 Sears p <.01 <.01 <.01 <.01 "Ideal"- z -6.70 -9.91 —lo.52 -ll.34 p <.01 <.01 <.01 <.01 . Federal aEx xcept where noted, first store in each comparison was rated higher by the social class in question b CValue of T with N > 25. Second store was rated higher. dProbabilities or true null hypotheses. 0 llO lowest. The findings also concur with Martineau's con— clusion that consumers, regardless of social class, distinguish between department stores on a socio- economic basis.1 The fact that consumers classify department stores in this way is important to store management. It implies that action taken to maintain or change the store image must be based on knowledge of the precise nature of the image. What is the image of the store in the eyes of consumers? What are the most important factors in this image? How can image maintenance or change be best accomplished? Martineau emphasizes the importance of the subtle impressions made by the store's advertising and interior decor.2 Management should also be concerned with the affect on the image of the type and variety of mer— chandise sold, the actions and attitudes of sales per— sonnel, the congeniality of the store, and the store's locational convenience. In all comparisons on merchandise suitability the subjects in each social class perceive statistically Significant differences among the test and "ideal" depart- ment stores. This finding is interesting in that although the total merchandise mix carried by each store is lMartineau, Motivation in Advertising, op. cit., pp. 173—175. ‘ 21bid. , Chapter XV. 111 different there are products which are common to each out— let. Does this mean that an item sold in Hudson's has different psychological qualities than the same item sold in Sears or Federal's? If this is so, and I sus— pect it is, the merchandising policies for identical products in different stores must be different. They must be designed to be congruent with what the consumer expects. Comparisons on the sales personnel dimension be— tween Hudson's and Sears and Hudson's and Federal's indi— cate that only the members of the lower classes perceive statistically significant differences. It may by hy- pothesized that this reflects the opinion held by many shoppers in the upper classes that the quality of sales personnel in the better department stores is declining. This points to a need for more emphasis on recruiting, selecting, training and compensation of sales personnel, especially in those stores in which the consumer expects high quality service. The results of the comparisons on store congeniality 1 show that, overall, consumers in each social class do per— ceive statistically significant differences between the stores. This, no doubt, reflects the differences in mer- chandise and sales personnel as they affect the “atmos— phere" in the store. An exception to this are members Of the lower class who perceive no difference between 112 the congeniality of Hudson's and Sears. It may be that such consumers have had little experience in these stores or that both stores are perceived of as being above them and so similar. One implication of this finding is that the atmosphere of Hudson's and Sears may discourage, by management intention or not, lower class shoppers. In general, the comparisons of locational convenience reveal that subjects in each social class perceive statis— tically significant differences among the test and "ideal" stores. However, as in the case of store congeniality, the members of lower class do not perceive a statistically significant difference between Hudson's and Sears. It appears that if the managements of these stores wish to appeal to lower class consumers they must discover unique advantages to offer to them. Comparison of Dimensions of Aggregate Department Store Images by Family Life Cycle Stage A summary of Wilcoxon T and probabilities associ— ated with the null hypotheses that there are no perceived differences 0n image dimensions among test and "ideal" stores for subjects classified by family life cycle stage is presented in Table l2. The results indicate that, overall, subjects grouped by family life cycle stages do perceive differences among the test and "ideal" 113 TABL12.--Summary of Wilcoxon T and probabilities associated with the null hypotheses by family life cycle stages for image dimensions between test and "ideal“ department D t t St a Family Life Cycle Stagese epar men ores 1 2 3 U 5 Merchandise Suitability liudsons- z: -3.1u -6.09 -8.ou b -5. at: —u .114 are p (.0 (.01 <.01 <.01 (.01 Hudsons- z —u.22 —10.68 —11.49 -7.79 —5 83 Federals p (.01 (.01 (.01 (.01 .01 “Ideal"- z -u.51 -9.97 -9.66 —5.85 -3 NH Hudsons p <.01 <.01 <.01 <.01 <.01 Sears- 2 —2.6L -8.u5 —9.05 —5.91 -u 75 Federal: p .0081 <.01 (.01 (.01 .01 "Idea1"- z -u.98 -ll.62 —ll.51 -8.20 -5 48 Sears p (.01 .01 .01 <.01 ( 01 "Ideal"- z -5.01 —12.07 -l2.01 -8.89 -6 50 Federals p <.01 (.01 (.01 (.01 <.01 Sales Personnel Hudsons- zb - .7h -0. M9 — .70 -l.h6 -3.02 Sears p° .081 .6212 .uauo .1uu2 <.01 Hudsons— z -1.BU -U.1U —5 68 -3.3U -2.8U Federals p .0658 <.01 <.01 <.01 <.01 "Idea1"- z -h.ll -9.56 B.U2 -N.98 -3.45 Hudsons p .01 <.01 <.01 <.01 <. Saars— z —l.2u -u.28 -5.l9 —3.2u -O. 59 Federals p .215 <.01 <.01 <.01.9522 “Idea1"~ z —u.7u -9.7u -8.19 —5.2u -5.61 Sears p .01 .01 (.01 (.01 (.01 "Ideal"- 2 -h.50 -ll.06 —lO.51 -6.96 -5.lB Federals p .01 (.01 <. <.Dl <.01 Store Congeniality Hudsons- z: —2.2 -3.95 —5 40 -3.81 —2.h7 p .n226 < n1<.01 <.01 .0126 Hudsons- z -2 59 - ii - :3 — .‘H —3.“l Federals p (.01 <.Jl <.01 '.Ol <.01 "Idea1"— z —3.33 —8.22 -7.V” -h 05 -3.90 Hudsons p .01 <.01 (.01 <.01 .01 Sea 2 «1.51 -3.7“ -5.Cl -3.58 —2.60 Federals p .1310 <.01 .31 «.01 (.01 "Idea1"- z -U.50 -9.32 -9.65 —b.0U -5.23 Sears p (.01 <.0l < 01 <.01 <.01 "Ideal“ z -u.uu 40.95 —l”."? 4.60 -5141 Federals p (.01 (.01 <.01 <.01 (.01 Locational Convenience Hudsons- z: -2.62 -u.2u -14.91 -2.53 -3.20 p <.01 <. <.01 .011u <_01 Hudsons- z -2.73 ~5.00 ~6.36 ~2.12 —h.09 Federals p (.01 <.01 <.01 .03u0 <.01 "Ideal"— z -u.u3 -10.5U —9.77 -7 be -5.11 Hudsons p (.01 (.01 (.01 (.01 (,01 Sears- 2 -0.97 -1.18 -3.2N -0 us -0.63 Federals p .3320 .2380 .01 .6h56 .5286 "Ideal"- z -u.9o —11.36 —11.13 -8.17 -5.61 Sear p <.01 (.01 <. <.01 <_01 "Ideal"- z -u.63 -11.38 -11.31 —8.23 -6.13 Federals p (.01 (.01 (.01 <.01 <_01 apt as noted, first store in each comparison was rated higher by the familya life cycle in question. bSecond store was rated higher. dProbability of true null hypotheses. °Va1ue of T with N > 25. 8Family Life Cycle Stages: 1 — Young, Single and Young, 2 - Yo ung. Married with children, 3 - Older, 4arried no children; Old Married, no children; 5 — Older, Single. Married with children; M 114 stores. In 84 of the 96 comparisons the differences are significant at the 5 percent level. Hudson's is rated higher than Sears in all cases but one and in all cases higher than Federal's. The "ideal" is rated higher than Hudson's, Sears and Federal's in all comparisons. Sears is rated higher than Federal's by all family life cycle stages. These results are parallel to the findings on social class and support the contention that the test stores are perceived in a hierarchy with Hudson's on top and Federal's on the bottom. These findings add to our knowledge of the com— munality of images of department stores. The implication of the results is that the classification of department stores on a quality continuum by consumers is a general phenomenon. This supports the notion of the "public image" described by Boulding.3 The results of the comparisons on the individual image dimensions held by consumers in various family life cycle stages are not as clear cut as those for social classes. Comparisons on the sales personnel dimension between Hudson's and Sears indicate that the differences are not statistically significant for subjects in life cyCle stages one through four. This is also true for 3Boulding, The Image, op. cit., p. M8. 115 the comparison of Hudson's and Federal's for members of stage one and for Sears and Federal's for members of stages one and five for sales personnel. This parallels the results of the comparisons of Hudson's and Sears' sales personnel by social class. It is further evidence that the consumer no longer feels that the traditional department store offers superior personal sales service. The fact that the older single consumers tend to rate Hudson's significantly higher may result from a "halo effect" from past experiences. In addition, there are no perceived differences be- tween Hudson's and Federal's on locational convenience by subjects in stage four and no perceived differences be— tween Sears and Federal's by subjects in stages one, two, four and five on locational convenience. These results may indicate that each of the stores has done an effective job in store location. Or, the findings may be inter- preted to mean that locational convenience, although important, is not a major factor in establishing the image of a department store. Comparison of Dimensions of Aggregate Department Store Images by Family Life Cycle Stages Within Social Classes Table 13 presents a summary of the Wilcoxon T and probabilities associated with the null hypotheses that there are no perceived differences for image dimensions ll6 8.. 8.. 8.. 8.. 8.. 8.. 8.. 8.. 8.. 8.. mz 8.. 8.. 8.. 8. 20 mo. 8.. 8.. 33 a mw.m- 004m- ho.:n 3.0- mm;- 0m.zr H.08- 00.0- 8m;- 3...: s San—ouch 88 8.8 88 8;. 88m 88 88 88 88 88 .H. 288?. 8.. 8.. 8.. 8.. 8.. 8.. Hr... mo. 8.. 8.. m2 8.. 8.. 8.. 8.. m2 8. 8.. 8.. AS a 8.:- 0H.N- 88- 8.? 88m- mmN- £8.- 88? mm...- H:..m- u 23m 8.8 8.3 88 8.m 88m 88 SN 8;. 8.: 88 .H. 3.803.. ma mo” m2 m8. m2 m2 8.. 8.. 8.. mo. m2 m2 8.. m: wz HHV 30 mo. m: 3.0 a S 8. HM N- 8H- 8.H- 87..- mN.m- 8.? 88- 8.8 mva 8 3238 8.8 88.1. 8.0 88.: 883 88 88 88H n88 .H. 5.8% .... mo. 8. 8.. 8. 8.. 8.. 8.. 8.. 8. m2 8.. 8.. 8.. m2 m2 8.. 8.. Ho 8 mm.H| >0.N- H0.Nu 0:.m- . .H.mn m0.m- wo.m- 08.0: mm.:- mm.m- u woomusm 8 mN 8.: 8.: 8; 8.8 8.3 8.: 8.8 88 H. 2885. 8.. 8H 8.. 8. m: L: 8.. 8.. m2 m2 m2 8.. we :0 m: .8. m2 Ho 8 . 8 N- N:.m- . KH- 8.N- (T:- i.N- 38- NmN- a 38.888 8 8 88H 8.8 8.2 88.3 8.8.3 8.N 88 nomém 984 .H. A588: 8.. ... mz m: mo. 82 E m 8. m: m2 m2 uz m: 8. m2 m2 m2 :0 d .. 3 o- 8.7 2.8- . S8- 88.H- 8.N- 888- nNNH- E8- " 38w m: n- 88H 8;... 8.8 88.2 8.8 8...? 8H 8.2 88.8 8.m a unsung: Hmccompom moanm 8. 8. 8. 8. 8. 8. 8.. 8.. 8.. 8.. 8.. 8.. 8.. 8.. 8. :0 mo. 8“. 8.. :0 d 8.:- m:.m- 88- 88- 88- NN.m- 8;.- 88- 88- NN m- . . N mHEMHam 88 88 88 88 88 88 88 88 8 o 8 8 .H. 3.8 8H. 8. 8. 8. 8. 8. 8. 8.. 8.. 8.. 8.. m2 8.. 8.. 8.. 8.. we Ad 8”. 8.. AC m I . ram 88- .2.:- mN.m- H:- 88- R...- JNH- 88- 8:- . 8H m- . . - H338 88 8.8 8H 88 8.8 88 8.: 8 o 8 o 8 8 a .. .. .. . . . a 8. 8. 8. 8. m2 m2 8.. 8.. 8.. 8.. oz 8.. 8 . 8 . mz H3 3 8 . 8 . “H0 u m . . . . . . . - . - 8.88.8 NN..- 8N T Z 1- N8 - 88- 8.:- 88- 8 .i- 8 m H: s . . - _ 88m 8.? 88 88 8.8 8.3 8H 8H 8 HH n.8 H .H 28w . . . . a m8. 8. H8. 8. 8. 8. 8.. 8.. 8.. mo. 2 8.. 8.. 8 . m8 2 oz 8 . 8 . 3.0 a neon . . . I on: TN: Nm.N- N .m HN.m- 8.8- 2.:- 58- .88.. em 0- MH m- 8N a . . - - m. m 88 88m 88 8.3 88 88H 8.8 8 HH 8 8 M :88? 8. mm. Mm. WWW 8. 8. 8.. “mm mm 8.. we 8.. “MM HHwnm- m2 HHV m2 mmnm- 8 . :0 n 3288 HN.- .- .s- ..- 8.- .- .- -. . . . . s z : 00.m 00.0 = 00.H 00.5 00.md 00.: 00 0 00 0 om z 00 m W nuanunm 8. m0. m2 m2 mo. mo. 8.. 8.. H0.m m2 mz Hmm Mmm mmnm mo. mo. 8.. “muw- 8 . A3 a 930m mfN- 8H.N- iH- $.H- m .N- 88- NH. - N . - . - . - . . . . . Leon : o 8;: 88m 8 8.: 88 8 m 8 N 8 o 8 S 8 H a u m mudaanoufizm omaocmnonoz m a m N H m n m N H m a m N H m a m N H mmwmum wHozo ouaq aaasmm Loud noun: nmzoq wanna: Lmaoq manna: D moumsHo Hmfiuom .noLOum acoEunmooo :Hmwoa: 0:0 umou Hmauow Canada mommum cacao mafia zHHEmw an mommnpoozz Has: on» :cozpmn macamcosun onEfi new mommsao Law: equHoommm mcfiuaafinmnona can a conceaaz mo hhuaaamll.ma m4mJ .vmuoc mhmnz ammuwm .coflpwmzu ca macho mhda mHflENu an hm£MMn Umumn mm: cowahmmEou .Hw>mm ”swuhma m>Hm mnu uw .mucwuwmncmHm vumeuwm f umumm mm; nemflmmufioo an mucum occumm .mm v E cpfl: B no wdamfiu n nuwm ad whOum umnamu pumOHhHumHm no: - mg on Hamam ecu : I flay Ho.. Ho.v Ho.v Ho.. #0.. Ho.. Ho. - . . - . - . - . 1 . 1 . H8 . mo 8.. H8. H8. m . me. Ho. Ho.. HHV a w: 3 mm m mm m m: w m. . wa.zl om.:: mm.m- r m>.m| aH.m| o AHV mo.m| N wamnmnmm H8.. H8.. . o m 88 OH 88.8 88.8 om.8 88.8 88.H 88.8 88.N 88.8 a -:HmmuH= Hm.m- 88.8- wm.m- Hm.m- Ho.. H8.. H8.. H8.. H8.. H8.. H8.. H8.. Ho.. Ho.. m8. mz Ho.. Ho.. Ho.. AHV a = m. . Hm.:l mo.8- Hm.wl mm.ml o:.mI Hm.ml u mhwww mz . , m2 0 m om mm 08.8 88.8 88.8 08.8 88.8 88.8 . 88.8 00.0 a I=Hmmnmg m».Hu MN.H- mMiH- m¢ - mz mz m2 mo. m2 m2 mg m: m2 m2 m2 HHV HHV mo.. mo. HHV a HH 8 . . amm.o- 88.H- Hm.o- mm.8- Nm.o- mm.m- N mHmpwUmm . 8m mm 888 8m 8m.mH 888.m 588.88 88.HH 88.8 oo.m 88.8N oo.m a -npaum ww.m- mm..- Ho”.- Ho”. Ho.. mo. Ho.. Ho.. Ho.. m2 m2 Ho.. Ho.. Ho.. Ho.. mz mz Ho.. Ho.. HHV a 8 8 a mo 0- . . m:- Nm.m- .38- 88- 3.8- 2;.- 8H.m- N 2325 Ho 88 m cm mH 88.m 88.8 88.8 8m.m om.HH oo.mH 88.8 a -.HmmoH= .. 82 82 82 82 n: ma 8. mo. 8.. m . . . . . . I . - . . 2 H8 H8 H8 . wz H mz H8 . H8 . AHV 8 mm m an 0 mm HI mo H1 . m:.cn ww.H- mH.m| r . mm.m- m~.=u A v Hm.8v N mHmpmnwm m2 88 mm om.8= 88.8 88.:H 88.88 88.8 88.H oo.m 8m.H oc.m a -m:0mo=m mz m2 mz mz mo. mz mz mo. mo. 7: -8. Ho.. . . . .. . - . - . - . o. H . Ho . m2 wz m8 Ho . .8 HHV a Nm 8 m8 0 Hm 8 mm Ha . 88.Ha mm.o- 8N.m- 88.m- wo.m- Hm.m- wN mummm am am 88.»~ 88.m 8m.m 88.8m 88.8 88.8 8m.m 88.8m oo.H as -m:omuam mucoficv>zoo am:o«unuoq Ho. H8. H8. Ho. Ho. m2 H8.. H8.. H8.. 88. m: -.. H8. H8. mo. H HH Ho.. Ho.. HHV a 8m;- m:- 8..? 8.8- :.m- 82- 88.8. 2 . 3w- Him- A 8 V m:- N 32%.... co.» om.m: oo.m 88.8 88.HH oo.H 88.8 88.8 88.m oo.H a -.HaocH: Ho” Ho” Ho” Ho. Ho. mo. Ho. H8.. H8.. m8. m2 H8.. H8.. H8.. mo. HHH m8. H8.. H8.. HHV 8 S 8- 8H m- 88 m- NE.- . $.N- R;- m=.m- :8- 88- 88...- N 238 on.” om.m~ om.= om.m 88.HN . 88.8 oo.m 88.: om.8H 88.8 a -gHaouH. He” no” Ho. wz wz m: wz m2 H8.. m2 w: m2 Ho.. mo. mz AH“ HHV H8.. mz HHV a 88 N- mw mu mN.- H8.8- 8:.H- 8H.H- mp.m- wm.m- mm.- 88.N- . N mHauoumm nom.hw 88.8m 88.» noo.HH om.mm 88.N 88.8 88.: 88.8N 8m.H a unnaow Ho” Ho. Ho. Ho. Ho. mz m2 H8.. H8.. m2 m2 Ho.. Ho.. Ho.. mz HHV mz H8.. H8.. HHV a mH =- mo.m- om.m- mm.=- 8m.H- 8H.m- N8.8- mm.=- mw.m- N8.N- N mcomuax 8m.8H 88.88 88.8H 8m.NH 88.88 om.8 oo.m 888.8N 8m.8N 88.8 a -=Hmw8H= mo. m8. m8. mo. wz mo. H8.. Ho.. Ho.. mo. m: Ho.. H8.. H8.. wz HHV m8. H8.. m8. HHV a 8m.N- 8M8- H=.- S8- 88- Him- 8...- .38- 85...- 38- N "838$ 88.8m om.=m oo.m 88.: oo.NH 88.8 88.8 88.H om.m~ oo.~ s -uzomuam m2 m2 m2 m2 mo. wz mo. m2 mo. 82 mz H8.. H8.. Ho.. mz AHV H8.. H8.. mz HHV a mm.8- mm.o- mm.o- H8.H- -.- 8~.H- NN.N- om.=- NH.m- NH.8- N «noun om.mH 88.8N 88.8H 88.8 88.8m om.HH oo.H 88.N om.NN oo.m a uncomuam huuauucowcoc okaym 118 among the test and "ideal" stores by subjects in family life cycle stages within social classes. The results indicate that consumers grouped by family life cycle stages within social classes do perceive differences among the test and "ideal" stores. Over two-thirds of the 480 comparisons of the test and "ideal" stores yield , differences that are significant at the 5 percent level. One-quarter of the 162 comparisons which did not yield statistically significant differences occurred in cases where the N was too small to estimate significance.“ In only twenty comparisons of the test stores by subjects in family stages is the second store in a com- parison rated higher than the first store. Thus, Hudson's receives higher ratings than Sears and Federal's, and Sears receives higher ratings than Federal's. This is further evidence that the test stores may be placed on a continuum from highest to lowest. The "ideal" store is rated higher than the test stores in all comparisons. These results agree with those found in comparisons among stores by social class and by family life cycle Y stage which show that department stores have distinct aggregate images. Department store executives should be aware of the image of their store and the images of Competing stores. The marketing strategist may define 11This occurred primarily in life cycle stages one and five in the upper class. 119 his role as exploiting his firm's perceived strengths and his competitor’s perceived weaknesses. In evaluating present and future store policies consideratiOn should be given to their effect on the store image. Merchandise suitability.——The comparisons among the test and "ideal" stores on merchandise suitability by respondents in life cycle stages within social classes indicate that perceived differences do exist. Almost 80 percent of the comparisons yield differences which are statistically significant. In all cases Hudson's again is rated higher than Sears and Federal's, and Sears is rated higher than Federal's, and "ideal" rated higher than the test stores. These differences in consumers' perceptions of the merchandise among test stores make it imperative that the marketing practices of the stores be different. For each department store the merchandise mix, prices, displays and promotions should be designed to be congruent with the consumers' image of the store's goods. Although no research evidence is available, it is hypothesized that marketing actions which diverge greatly from the consumers' image of a department store might tend to damage the store's image by confusing consumers. Heidingsfield warns that unclear imagery may result in "5 "no significant market penetration. 5Heidingsfield, "Building the Image," op. cit., 120 Sales personnel.——Consumers classified by family stages within social classes perceive differences on the sales personnel dimension among the test and "ideal" stores. Approximately 64 percent of the comparisons on sales personnel result in differences which are signifi- cant at the 5 percent level. However, in only A of the 20 comparisons between Hudson's and Sears are the per— ceived differences on sales personnel statistically significant. Also, in comparisons between Hudson's and Federal's and Sears and Federal's only about one—half the comparisons yield results that are statistically signifi— cant. However, in over 80 percent of the comparisons among the "ideal" and the test stores the differences are significant at the 5 percent level. Consumers apparently perceive the test stores as lacking in the sales personnel dimension compared to their "ideal." In comparisons of sales personnel, where the differ— ences are statistically significant, consumers rate the first store in each comparison higher. Again, overall, Hudson's was rated higher than Sears and Federal's, and ’ ' Sears was rated higher than Federal's, and the "ideal" rated higher than the test stores. These results parallel those found in comparisons Of sales personnel among the stores by social class and by family life cycle. Consumers, especially at the higher social class levels and in the younger family 121 stages do not perceive of differences among the test stores' sales personnel. From the consumers' point of view, none of the stores has a competitive advantage in this area. It may be assumed, because the traditional depart- ment store was formerly recogniZed for its superior sales service, that the relative level of personal sales has declined. The members of the higher social classes may be expecting better quality personal service than they are receiving. In addition, consumers in the younger family stages may be more concerned with other aspects of the buying process, such as merchandise selection and price, and so pay little attention to sales personnel. Store congeniality.——The comparisons among the test and "ideal" stores reveal that consumers in family life cycle stages within social classes do perceive differences in store congeniality. About 65 percent of the compari— sons yield differences that are significant at the 5 per- cent level. However, there are some exceptions to this trend. The first is the comparison between Hudson's and Sears by the lower class in which subjects in various life Cycle stages do not perceive differences that are statistically significant. The second exception is in the comparisons of the test and "ideal" stores by 122 respondents in the first and fifth life cycle stages in the upper middle, lower middle and lower classes in which only half of the comparisons yield results that are significant at the 5 percent level. It may be that members of the lower classes and the younger stages have little experience on which to discriminate between Hudson's and Sears. Younger and lower class shoppers may not have developed habitual shopping patterns and may be more amenable to persuasion. This suggests that a department store which directs marketing effort towards these consumers may benefit by their patronage as they move into later stages of the life cycle and/or move up the socio—economic scale. Locational convenience.——Comparisons among the test and "ideal" stores on locational convenience by subjects in life cycle stages in social classes yield mixed results. Overall, about 60 percent of the comparisons between stores yield differences which are significant at the 5 percent level. For lower class consumers at various stages of the life cycle, comparisons of Hudson's—Sears, Hudson's—Federal's and Sears—Federal's yield results which are not statistically significant. This agrees with the results of the comparisons among stores by social Class. It may be that for the lower class consumer locational convenience is not a meaningful dimension on Which to judge a department store. 123 Comparisons between Sears and Federal‘s by family life cycle stages in the upper middle and lower middle classes also yield results that are not statistically significant. This is in contrast to the results of the comparisons by social class which showed significant differences between these two stores. It may be hy- pothesized that the locations of the Sears and Federal's stores do not yield either company a competitive ad— vantage with middle class consumers. However, subjects in the upper and upper middle do perceive differences between Hudson's and Sears on 10— cational convenience. Upper and upper middle class con— sumers may perceive Hudson's as significantly better than Sears on location for two reasons. One, the Hudson's downtown store is close to their business offices. And two, Hudson's major branch stores are located in higher class neighborhoods while Sears' stores, with few ex— ceptions, tend to be located in middle and lower class areas . CHAPTER VI DEPARTMENT STORE IMAGE: SOCIAL CLASS AND FAMILY LIFE CYCLE FACTORS This chapter presents the results of the affects of social class, family life cycle and social classes at various life cycle stages on consumers' perceptions of the test and ”ideal" department stores. Social Class Affect Table 14 summarizes the Mann—Whitney U scores and the probabilities associated with the null hypotheses that there are no perceived differences on the four image di— mensions of the test and "ideal" stores among social classes. The majority of comparisons (50 of 96) among social classes do not yield differences which are statistically significant. But, there are differences among the images held by members of social classes which can only be clearly seen in comparisons of widely separ— ated classes. Over 65 percent of the comparisons in which there is a significant probability of a difference between social classes resulted in cases where the classes COmpared were not adjacent to one another. 124 TABLE lll.--Sumrnary of Mann—Whitney U and probabilities - hypotheses for image dimensions of tee associated With the null 125 "ideal" department stores between social classes . Social Classesa Department Stores Hudsons Sears Federals "Ideal" Merchandise Suitability Upper— 2c -2 29 1 5 d - - . 7 -o.65 -o.19 Upper Middle p .05 NS NS NS Upper- z -1 1n -2 3n” 2 62b . . - . -2.U8 Lower Middle p NS .05 .05 .05 Upper- z —1.99 -3.67b 41.52" -u.u1 Lower p .05 .01 <.01 <.01 Upper Middle- 2 —1.67b -u.62b -u.oab -2_56 Lower Middle p NS <.01 <.01 .05 Upper Middle- 2 —o.65b -6 In" .6.75b -5.10 Lower p NS <.01 <.01 <.01 Lower Middle- z -1.18 -1.57b .3.02b -2 73 Lower p NS NS <.01 .01 Sales Personnel [pper- 23 —1.80 -1.21 —1.01 —o.22 Upper Middle p , NS NS 3 NS Upper- z —o.18 —o.36b .0.05 —1.37 Lower Middle p NS NS NS NS Upper— z —1.2ob -o.65b -1.95b -1.10 Lower p NS NS .05 NS Upper Middle- 2 -2.11b —1.82b -1.19 -1.92 Lower Middle p . NS Ns NS Upper Middle- 2 -3.73° —2.21b -3.29b -l.62 Lower p <.01 .05 <.01 NS b b b p Lower Middle— 2 -1.82 -O.52 —2 ”6 —O.39 Lower p NS NS .05 NS Store Congeniality Upper- zc -0.69 -o.uo —o.68b -o.25 Upper Middle pd NS NS NS us Upper— z -O.D8 -l.69b -i "'3“ -2.33 Lower Middle p NS NS <.01 .05b pre._ z -o_53 .3.38b -5.17“ —o.01 Lower p NS < .01 < .01 NS Upper Middle- 2 .o.35b -2.59b ~u.1o -2,5u Lower Middle p NS .05 <.01b .05b Upper Middle— 2 -o.36 _u.63b —6-12 -0‘32 Lower p ns <.01 <.01b NS 1 Lower Middle— 2 -o.12b —2.05b ~2-63 -2 93' ower p NS . 05 . 05 . 01 Locational Convenience 6 2b 1 07 Upper— zd ~1.l7 '1-57 '1-3 ' ' Upper Middle p NS NS NS b NS Upper— z -0,37 -2.5ub -u.aa -3.eu Lower Middle p NS .05b <-01b <-01 Upper- z .1.01 -3.59 6.06 -3-“6 Lower p NS <.01 <.le <.01 Upper Middle- 2 -o.97b -“-93 —u.a9 '2 63 Lower Middle p NS <.01b <.01 .05 Upper Middle- 2 -0.33b -6~23 ‘5-00 '2'u5 Lower p NS <.01b <.01 .05 Lower Middle- 2 -o.76 -1.33 -0-05 -°-22 Lower p NS NS NS NS aFirst social class i where noted bSecond social class in comparison rate 6The value of U with N2 > 20. n each comparison rated store 6 store in dProbabilities in question higher, except question higher. of true null hypotheses. 126 It is interesting to observe the direction of the preferences shown by the comparisons among social classes. In a majority of the comparisons in which the differences were statistically significant the lower class in the comparison rated the store in question higher. This phenomenon was especially prevalent in comparisons of Sears and Federal's. It appears that the relatively lower classes are less critical of these two stores while the upper classes tend to be less critical of Hudson's and tend to set a higher standard for the "ideal" store. This may be interpreted to mean that consumers in the lower classes prefer Sears and Federal's while con- sumers in the upper classes prefer Hudson's. If this is true, it tends to support Martineau's hypothesis that a department store cannot be all things to all people. He writes, ". . . there is no such thing as a store image with equal appeal . . . for all social classes. . . ."1 However, the results do not suggest that depart- ment store managers act to limit the appeal or services offered to only one group. Although members of one social class may be most strongly attracted to a particular department store, there are members of other social classes who also prefer that store. It may be possible for a department store to create different types of branches to appeal to different social lMartineau, "The Personality of a Department Store," op. cit., p. 50. 127 class groups. Physical location, store design, mer— chandise selection and personnel could be planned to meet the needs of the social class of the customer population. MerchandiSe'Suitability Consumers in social classes appear to have dis— tinctive images of the suitability of merchandise in the test stores and have a differentiable image of an "ideal" store on this dimension. AlmoSt 70 percent of the com— parisons among classes on merchandise yield statistically significant differences. As in the analysis directly above, two—thirds of the significant differences result from comparisons of social classes not adjacent. The direction of the preferences also fits the general pattern in which over one—half of the comparisons show the lower class rates the store in question higher. Again this occurs almost totally with regard to Sears and Federal's while the relatively higher classes are more favorable to Hudson's and present a higher hypothetical "ideal." It is assumed that Sears and Federal's offer merchandise which fits the desires and the budgets of lower class consumers while Hudson's merchandise satisfies higher class consumers. Although there are some differences among members Of various social classes in their taste for merchandise these differences are only distinct between non-adjacent 128 classes. This implies that the market for a particular classification of merchandise is fairly broad. A depart- ment store with a limited range of merchandise can appeal to consumers in more than one social class. Sales Personnel There does not appear to be any perceived differ- ence on the sales personnel dimension of the test and "ideal" stores among social classes. Only one—quarter of the comparisons result in differences which are statisti— cally significant. Again 65 percent of the differences which are statistically significant come from comparisons of classes not adjacent. All of the comparisons which have significant differences indicate that the lower class rates the store in question higher. As in the pre- vious analysis the lower classes prefer Sears and Federal's but, in addition, also rate Hudson's higher than did the relatively higher classes. It seems that the higher classes perceive something lacking in this aspect of Hudson's operations. The higher ratings given by members of the lower . classes may indicate lower expectations in personal ser— vice. It is also possible that the sales clerks are themselves members of lower classes and so better able to serve the needs of lower class consumers. This is a critical area for store management. Harris contends that people base their impressions of a store primarily 129 on the attitude of the sales personnel.2 This contention should be researched. The results may indicate how a department store can project a better image by improving its sales people. Better training, compensation and supervision may result in a better aggregate image. Store Congeniality Consumers in social class groups appear to have distinguishable images of store congeniality for the test and "ideal" stores. Half of the comparisons among social classes on store congeniality result in differences which are statistically significant. Almost 60 percent of the differences which are statistically significant come from comparisons of non-adjacent social classes. Two—thirds of the comparisons show that the relatively lower classes give higher ratings than higher classes for the store in question. Again, the lower classes perceive Sears and Federal’s more favorably. This was true in every compari— son with the lower class. It may be that consumers in the lower class need a greater measure of store congeniality to feel comfortable while shopping in a department store. This conjecture agrees with Martineau's contention that the members of the lower classes are insecure in shopping situations.3 2Harris, Buyers Market, op. cit., p. 103. 3Martineau, "Social Class and Spending Behavior,” Op. cit., p. 129. i1 130 Management may be able to make lower class shoppers more comfortable in the store environment. Lower class con— sumers will feel more secure in a store where they know they will receive a pleasant reception from the clerks and where they can afford to purchase at least the low end of the merchandise selection. They will feel more comfortable in a store which offers credit privileges to responsible lower class shoppers. Locational Convenience Consumers in social class groups appear to have dis— tinguishable images of locational convenience for Sears, Federal's and "ideal" but not for Hudson's. In addition, comparisons of upper and upper middle, and lower and lower middle reveal no statistically significant differ— ences. Overall, 50 percent of the comparisons yield differences that are statistically significant and 85 percent of these (compared to 65 percent overall) are from social classes not adjacent. As was the case overall, 50 percent of the comparisons indicate higher ratings by the lower classes for the store in question. Again the relatively lower classes give Sears and Federal's higher ratings and the relatively higher classes present a higher "ideal." Consumers in all classes apparently have a similar View of the locational convenience of Hudson's. This may reflect Hudson's attempt to cover a broad spectrum 131 of geographic and economic markets. Lower class con— sumers favor the locational convenience of Sears and Federal's which is an indication of the success of their location policies. Consumer perceptions of the appro— priateness of a location may be a useful guide in new store planning. Family Life Cycle Affect Table l5 summarizes the Mann—Whitney U scores and the probabilities associated with the null hypotheses that there are no perceived differences for the image dimensions of the test and ”ideal" stores among family life cycle stages. It appears that consumers grouped as to family life cycle do not perceive differences in the test and ”ideal" department stores. Of the 160 comparisons on the test and "ideal" stores only thirty, or less than 20 per- cent, result in differences significant at the 5 percent level. Seventy percent of the comparisons where the differences are statistically significant are between non— adjacent family life cycle stages. In those instances where differences are statistically significant, over one-half of the comparisons show that con— sumers in the second life cycle stage in a comparison prefer the store in question. Subjects in the older family stage in a comparison rate Hudson's, Sears and ”ideal" higher while consumers in the younger stage prefer Federal's, 132 TABLE l5.—-Summary of Mann—Whitney U and probabilities associated with the null hypotheses for image dimensions of test and "ideal" department stores between family life cycle stages. Family Life Department Stores a b Cycle Stages ’ Hudsons Sears Federals "Ideal" Merchandise Suitability 1-2 z: —1.08 —1.72 —1.02 —1.52 p NS NS NS NS 1-3 2 —l.80 —1.15 —0.17° —1.80 p NS NS NS NS 1—u z —2.22 —1.91 —2.00 , —0.u1C p .05 NS .05 NS 1-5 2 -1.50 —1.u5 -1.31 -0.56c ,p NS NS NS NS 2-3 z —1.18 -0.95C —2.1uc —o.u2 p NS NS .05 NS 2_u z _1 90 —0.56 —l.81 -2.73c p NS NS NS <.01 2-5 2 —o.81 —0.05 -o.73 -2.26c p NS NS NS .05 3—A z —o.96 —1.35 - —3.u3 —3.38c p NS NS <.01 <.01 3-5 2 —o.o6 —o.7o —2.13 —2.710 p NS NS .05 <.01 H-5 z —0.590 —o.3oc —o.69C —0.29c p NS NS NS NS Sales Personnel 1-2 z: —1.1u° —1.97° -0.95C —o.18C p NS .05 NS NS 1-3 z 4.98C —2.63c —0.79c —o.o3 p .05 <.01 NS NS 1~4 2 1.88° -2.66c -1.72C —1.72 ‘ p NS <.01 NS NS 1-5 z 2.72C -2.370 -2.080 —o.21c p <.01 .05 .05 NS 2-3 z ; 1.67C -l.23C —o.17 —o.25 p NS NS NS NS I 2—u z —1.2Ac —0.88° -1 50C -3 19 p NS NS NS <.01 2-5 z —2.suc —o.73° —2.1uc —o.05c p <.01 NS .05 NS 3-'l 2 —o.18 —0.08c —1.53C -3.00 p NS NS NS <.01 3—5 2 -1_790 —o.06c 2.15c —o.22C p NS NS 05 NS ' 4-5 z -1.67° —o.09°* —1.11c —2.28°_ p NS NS NS .05 TABLE 15.-—Continued. Family Life Department Stores ab CV01e Stages ’ Hudsons Sears Federals "Ideal" Store Congeniality 1—2 2: —o.36 ~o.55 —o.07 —o.99C p NS NS Ns Ns 1-3 z .0.70 -o.52 —o.52C -0.91 p NS Ns -NS NS 1—A z -1.au —1.57 —o.99 —1.22 p NS NS NS Ns 1-5 2 —1.u9 —1.32 —1.01 —o.59 p NS NS NS NS 2-3 2 —o.51 —0.oo —1.01° -O.33 p NS NS NS NS 2_u z —2.27 —l.6u —1.61 —o.12 p .05 Ns NS NS 2—5 2 —1.45 —1 28 —l.u2 —o.uu p NS NS NS Ns 3—u z —1.90 -1.61 —2.28 —o.u8 p NS NS .05 Ns 3-5 2 -1.1u -l.2l ' -l.98 -O.2U p NS NS .05 NS A-S z -0.5oc —0.05 -0.22 —o.55 p Ns NS Ns NS Locational Convenience 1—2 z: -l.32: —0.77° -0.u7c —o.81c p NS NS NS NS 1-3 z —o.5gc —o.69° —1.12c -0.71c p Ns Ns NS NS 1-4 z -0.03 —0 63 —0.91 —1.66C p NS NS Ns NS .1-5 2 _u.uu —0.<38U —0.35c —1.13c p HS NS NS NS 2-3 z _1.27 —o.16C -1.Ulc —o.29 1 p NS NS NS NS 2-4 z —1.65 —l.83 —2.23 _1,u7c p NS NS .05 NS 2-5 2 -1,75 —o.11C ~0.09° -o.5 C W p NS NS NS NS .1 3-A z _o,5o -2.00 —3.26 -1.80C ; p Ns .05 <.01 NS I 3-5 z —o.81 -o.03C —O.88 —0.82c p NS NS NS NS 4-5 2 -0.15 _1.uic -1.77c —o.51 p Ns NS Ns NS aFamily life cycle stages: 1 — Young, single & Young, married no children; 2 — Young, married with children; 3 — Older, married with children; A - Older, married no children; 5 — Older, single. bFirst family life cycle in each comparison rated store in question higher, except where noted. cSecond family life cycle stage rated store in question higher. dThe value of U with n2 > 20. eProbabilities of true null hypotheses. I l34 It appears that Federal's has a stronger appeal to con- sumers in the younger stages in the life cycle. The lack of a family life cycle affect in the re— sults may simply indicate that attitudes about depart— ment stores are not greatly influenced by family stage. This conclusion is in opposition to Collazzo's conten— tion that consumer attitudes are influenced by their stage in the life cycle.4 On the other hand, these findings may have come about because the definitions of the stages used in this study were too narrow to show differences. Support for this is indicated by the fact that 70 percent of the com— parisons yielding statistically significant differences were between non—adjacent stages. Data on the preferences of members of family stages for particular department stores is inconclusive but may indicate tendencies. Younger families may tend to rate Federal's higher because it satisfies their needs within their budget restrictions. If this is true, Federal's management may be able to capitalize on this by slanting their merchandising and promotional activities toward this younger group. In the same way Sears and Hudson‘s may desire to concentrate their attentions on somewhat more mature family stages while attempting a new marketing approach to the younger families. ”Collazzo, Consumer Attitudes and Frustrations in Shopping, op. cit., p. l. 135 Merchandise Suitability There does not seem to be any strong evidence of perceived differences of merchandise suitability of the test and "ideal” stores among consumers grouped by family life cycle stage. Only one-quarter of the com— parisons among family life cycle stages on merchandise suitability yield statistically significant differences. Further, 60 percent of the comparisons in which the differences were significant were between non-adjacent life cycle stages. Of the nine comparisons on merchandise suitability which yield significant differences, over 50 percent show that consumers in the relatively older stages rate the stores higher. Almost all of these higher ratings by consumers in the older stages are on the "ideal" while consumers in the younger stages rate Federal's higher. Apparently Hudson's and Sears are presenting their merchandise equally well to consumers at all stages of the family life cycle. This, no doubt, reflects the breadth of type and quality of merchandise available in these stores. The preference of the younger families for the merchandise in Federal's may result from narrower and less expensive lines. y Sales Personnel The results indicate that consumers do not perceive differences on sales personnel for the test and "ideal" stores among subjects in family life cycle stages. Approximately 30 percent of the comparisons of sales personnel among consumers in family life cycle stages ‘1 yield differences which are statistically significant. Of these comparisons 77 percent are between non—adjacent family stages. Almost 75 percent of the comparisons yielding signifi- cant differences on sales personnel of the test stores among family stages indicate that the second stage in a comparison rates the store in question higher. This may indicate that the activities of salesclerks are more important to the relatively older life cycle stages. As was the case with merchandise suitability, there appear to be some differ— ences in the perception of sales personnel among the younger and older stages. Sales personnel should be trained to be sensitive to the needs of consumers in the more mature life stages. Departments which cater to these shoppers should be staffed with clerks who are best able to relate to the older shopper. Store Congeniality There are no apparent differences in perception of store congeniality of test and "ideal" stores among 137 consumers in life cycle stages. Seven and one—half percent of the comparisons among life cycle stages on store congeniality of the test and "ideal" stores re— sult in differences which are statistically significant. Of the three statistically significant differences two are between non—adjacent family life cycles. All three of these differences show that subjects in the rela— tively younger stage rate the stores in question higher. These findings indicate that consumers in various family stages hold similar views of the congeniality of the test stores. This may reflect the diversity of the shopping areas within the stores but it seems unlikely that the atmosphere of a store would be equally attractive to members of old and young family stages. This con- jecture is supported by the evidence that, overall, the younger families tend to prefer Federal's and the older families tend to prefer Sears and Hudson's. Locational Convenience The findings indicate no perceived differences in locational convenience among subjects in stages of the life cycle for the test and "ideal" stores. The three Comparisons yielding statistically significant differ— ences come from non—adjacent life cycle stages, and again show that consumers in the relatively younger family stages rate the test stores higher. It might be concluded that the test stores have been successful or 138 unsuccessful in locating their stores so all family stages rate each of them equally. Social Class——Family Life Cycle'Affect Table 16 summarizes the Mann—Whitney U scores and the probabilities associated with the null hypotheses that there are no perceived differences in the test and "ideal" stores among consumers in family life cycle stages in different social classes. The null hypothesis is supported. There is no indication that there are perceived differences in the test and "ideal" stores among subjects in family life cycle stages in various social classes. There does appear to be a tendency for differences to exist between stages in non—adjacent social class, especially in life cycle stages two, three and four. Less than 20 percent of the HBO comparisons of the test and "ideal” stores yield differences which are statistically significant. Of the comparisons which are significant at the 5 perCent level, 60 percent result from comparisons of family stages in non-adjacent social classes. In addition, 90 percent of these comparisons are in life Cycle stages two, three and four. This is not consistent with the results of the affect of social class reported in the first section of this chapter. Those results indicated that comparisons among social classes did result in different images of 9 3 l mm -mmumn Mn .5... .. H loandl n: In .. I I . a .. . ...... mm when .w Ima“ p... "an” . mz mm. . . onoa.m mz m:.nI Mfl owwuwn M mz mo.wn mz =:.ol oc.em mz oo.mw N Ioflwwmw mz 0317 mo” usm.ml an owa.a| mz . m a . hosed mz m>.0I mo 0mm.NI o v mom.ml mo. 0w:.ol m wz . mz mo.HI mz semi? . oHa.NI 8 mm m mz . I mo mH NI e mz om.H. :o o mz wen. I m nosed mz 3.3. . nz 3;... mz o . H om.~m N I353: mz omN.0I MW. m:.H om.m: m2 ejmm m A human. . I I u . mu MM.MI mz Mzm.HI mwa prHHI mn 0H.OI om do m wz m. wz oHo.oI mz p =.NH mo. ma.HI : oHooHs a HH m2 oo.om m2 omm 0 mo. www.ml m hozoq mz p36- mz mm 3.2 mz e 3.3 m Iowwwwe mz . I .HI . a we me. me. Need. mz were- mz 25. H oz 2.5. mz pee T 2 . m mz me.HI o . m2 ee.eI m: were I mz pom.Hm mu ese 0. o mz www.3I mm oMe.OI m o. a . I mz om. m ma mz oo.mH m2 0 h a om.NH N nozoq mz no.0I m m wz woo.flm mz . w H Inmama wz mm.o| m2 mmz.HI wz mm.0I mom mm m2 om.>m oz 54- mm mm- m2 .34. Me 2.7 m mz . . I o . I z Hm.HI moo ma m2 om.w NH mm a mo. IHN.NI m UHUUHS mz m. I om.mH m: a oo.m m posed m2 OH.OI oo mm Ho. oom.> wz n m H Incas: . m2 em. - exam , . mu omw.wm mo. omm.m- m: eH.oI p mm n..HI oo mN m mz one we .85. my WWW me. RI m p m2 oo.m wz m2 Hp.eI oHeeHz moc.HH m2 o o .p m some: m o H lemons Hmccowsom mmHmm m2 mm.oI 0 . mz Hm.HI mu oeo.HI w: ms.HI . oz 3.? pg 0. a: pee..- me puma- mo. me.mI mme pMH.HI m pmm.mI m: eme.eI m m2 om.mm o om:.mI m2 mm.HI m2 mm.HI m b m2 oo.mm m2 o mo. 0:.NI posed mz . p oo.me m2 . m IpHoeH: omm oI mz . I mom mm H m2 HH.NI . :m o m: m . I Lmzoq Ho.v s=.mI Ho.v omm.mI Ho.v omm.m- m2 omm.0I m a. s..- an WM” 8.. Ms..- me .3..- . wz 00.2: wz o oo.m: mu om:.ml ma MN.WH m hosoq m2 om.wm m2 o oem.me me o em.ee m Immemwe 2 H . I p . sea . e e 8.. «WWI we. pmHT e mme mam- em 3 m: 8.: m me ma? 0 peed- 3.. .2. I m: 2.? p I mz oo.mH m2 ome.HI m2 pem.mI mz mee.HI m oHepez m2 om.mm m2 m . Ho.v www.ml N I hoxoq mz Hm.eI mz mm poo NH mz oom.NH H omwmwm mz m . I . .o- . I mz em.MI Ho.V Mom.mI mma mw.MI m2 Hm.0I m Ho.v mo.mI Ho.V omo.mI m2 owe.HI m2 em.HI : mz . mo o::.mI me. p . m2 om.HI 00 MH m2 om.mH m2 owe NI m2 mm.oI m mz oo. ooc.OH m2 oo.m m poxoq m2 mm.0I Hm we ooo.m: mz . m pods: wz mm.HI mo mom.mI mz ms.HI om Hm m2 oo.m: m m: :w.HI mm cmwnwl mz MH:.HI Mn mm.HI : m2 . I . I .a.OI .: e. . .3 a... .e. e m as“. mz . o mz 00.:N I m2 me.0I om mm m2 ooo.mH mz . H some: mz o>:.HI vz mammal mz mo.ol 00 :N mz oo.NN m mz MHHHI mz 0H HI Ho; ooo.mu m: .327 = m2 . m2 pHH.0I mz H . I mz :m.HI m 00 0H wz oo.s o H o mz :m.HI oHpon mz om.m mz .. N poem: 0 wow m H lemma: soHHHQMszm omHocmnopoz . a N Ham 3 a u a a N s d = UH: afimswuom mpmmm o nu ma m:0muzm Umaomo oqu muoHo hHHsdh HmHuom monouw ucoEuLmnoD .mouOuu uses mews m I phenom Has a macho MMHH hHHEdu Coozpun mumonmwmhm:MHwwownwonmmmdmcoan ow unawaoomm mEH sou mmmeHo HmHoOm :Hnawz m meuaHHnmnopa use 0 hoszsslccmz no mnmEESmII.wH wqu+a .ndmwnuonh: HHsc can» no noauHHHnmnoLmu poo.mm oo.mm 00.nm oo.mm om.0m om.mm oe.eH eeo.am poe.HH 1’40 mm.mI Mem.HI oo.OI pew HI me.HI omH.OI He.e- me.eI peo.~e oem.mm em.mo oe.em oem.mH oe.em 00.0H 00.:N om.HH mm.NI Ms..- mN.HI oem OI avaaaucowcao chaum .Ho>oH unnamed m on» ad ucaoau«:MHu no: I mz .nmnwun :oaomosv :« macaw roams :Omnpdnsoo :H momma Hwauom uncomm :u wanna «Home uudn hudsmmm .o~ . m: eon a do esHp>e 0m.mw oo.mm oo.Hw 00.hm ooo.em m00.ow wom.0H 00m.:m .em.m 00.00 00.0: rem.ee oo.mm 00.0H ooo.mm oo.MH poo.e~ mom.HH N v oo.m¢ 0m.Hm om.mh om.hm 0m.ma oo.mm 00.5 om.nN pom.ee om.o= oo.mh 00.>~ ooo.- 00.mm om.~H ooo.Hm 000.0H mea'ln HNm=m HNMJIA Hmmawn r-INM:an «mma'ln riNrn=wn H(Vfl1#ln Hcvnwa-m Hrva\= m riwffia-m ~1N'“='m .umuo: when: unwoxu .nwnma: soaumosv :« onouw copes :omHAmnEoo some :a mmMHo HmHoou umLHH :d uwmum oHozo mMHH hmfiaem U I I .e eon a do orHose IwHuUHZ Loan: Amzoq lawman educdx hwzoJ hosed IwHQUdS nosoq nosed Ioauvaz some: meUH: solo; lawns: UHVUHZ .r 1+. 1A1 the test stores. It appears that there is some com— munality of department store images within each family life cycle stage. However, as is shown in the section on family life cycle affect, there are few significant differences among family stages. Why the social class affect disappears when comparisons are made by family stage among social classes is not known. The influence of family life cycle on consumers' attitudes bears further research. For both Hudson's and "ideal," only eleven of the 120 comparisons among stages in different classes result in differences which are statistically significant. About one—quarter of the comparisons on Sears and over 30 percent of the comparisons on Federal‘s yield differ— ences which are statistically significant. There is some indication that the family life cycle—social class affect is greater for Sears and Federal's than it is for Hudson's and "ideal." These findings seem to indicate that Hudson's has a broader appeal than do Sears and Federal's. The variety in merchandising policies seen in Hudson's main store, major branches and budget branches may account for their high ratings across the various family stages and social classes. These results may also be inter- preted as an indication of the success of Hudson's attempt to serve many markets. This appears to refute 142 Heidingsfield's contention that if a store attempts to serve a very broad segment of the market its, "Image will be contradictory and confusing."5 About 90 percent of the comparisons which result in significant differences indicate that the subjects in family life cycles in the relatively lower classes rate the store in question higher. Subjects in family stages in the relatively higher social classes tend to present a higher "ideal." Members of the lower classes apparently have lower expectations and so rate the stores higher than do the members of the upper classes. This seems to support the notion of using varying qualities of merchandising, sales personnel and furnishings in different departments de— pending on the anticipated clientele. Merchandise Suitability There is no indication that there are perceived differences in the test and "ideal" stores on merchan— dise suitability among respondents in family stages in different social classes. Approximately 20 percent of the comparisons among consumers in family stages in different social classes result in differences which are statistically significant. About 60 percent of these comparisons come from stages in non—adjacent 5Heidingsfield, "Building the Image——An Essential Marketing Stratagem," op. cit. p. 139. I: 1A3 social classes. Over 90 percent of the differences which are statistically significant come from compari- sons of stages two, three and four. The direction of the ratings fits the overall pattern which indicates that consumers in relatively lower classes tend to favor Hudson's, Sears and Federal's while subjects in relatively higher classes tend to have a higher "ideal." Almost 80 percent of the comparisons which have significant differences on merchandise suit— ability show that consumers in the life cycle stages in the relatively lower classes rate the store in question higher. These results are not consistent with those in the first section of this chapter in which it was found that there were differentiable images on merchandise suitability among social classes. It is assumed from the research evi- dence that there is a social class affect but that it is somehow obscured when the social classes are divided into family life cycle stages. Why this occurs should be studied. 1 I Similar to the findings on social class, there tend to be more differences among members of family stages in non-adjacent than adjacent social classes. This sug— gests that a stores' merchandise will appeal to members Of adjacent social classes. It supports the contention 144 made in Chapter V that a department store can attract shoppers from a number of social classes. Sales Personnel Similar to the findings on merchandise suitability there is no evidence of perceived differences of the test and "ideal" stores on sales personnel among con— sumers in life cycle stages in different social classes. Slightly more than lO percent of the comparisons on sales personnel among subjects in family stages in different classes result in differences which are statistically significant. Of these, 70 percent are be— tween family life cycle stages in non—adjacent social classes. Almost 80 percent of the statistically signifi— cant differences result from comparisons in family stages three and four. The direction of the ratings on sales perSOnnel is different from the normal pattern regarding the "ideal." For the "ideal” store no statistically significant differ~ ences result from the comparisons and thus no direction can be assigned to the ratings. The results of the com— parisons on Hudson's, Sears and Federal's conform to the general pattern which shows respondents in the family Stages in the relatively lower classes favor these stores. 1A5 These results agree with those found in the analyses of the social class affect and the family life cycle af— fect reported above. The lack of differentiability among social classes of the sales services offered suggests that management attention be directed in this area. The higher ratings given by consumers in the lower classes may indicate that personal selling in these stores is of a quality which satisfies only lower class consumers. Members of the upper classes may have expectations which cannot be satisfied under present conditions. This important contention requires an answer. Store Congeniality There is no indication that there are perceived differences of the test and "ideal" scores on store con— geniality among subjects in the family life cycle stages in different social classes. About 18 percent of the comparisons on store congeniality among life cycles in different social classes yield differences which are statistically significant. Of these, 65 percent are be— tween family stages in non—adjacent social classes. Al— mOSt 90 percent of the differences which are statisti- cally significant come from comparisons on stages three, four and five. Similar to the findings on sales personnel, the direction of the ratings shows that the subjects in the family stages in the relatively lower classes rate 146 Hudson's, Sears and Federal's higher on store congeniality. However, in the only comparison on the "ideal" which re— sults in a significant difference, the subjects in the relatively lower class presented a higher "ideal." This lack of differentiability on store congeniality may be interpreted in a number of ways. It may mean that the atmosphere of a store appeals equally to members of family stages in each social class. Or, it may imply disinterest on the part of consumers. Locational Convenience As was found in comparisons on the other image di- mensions there is no indication that there are perceived differences in the test and "ideal" stores on locational convenience among subjects in family life cycle stages in different social classes. Twenty percent of the com— parisons on locational convenience among respondents in family life cycle stages in different classes result in differences which are statistically significant. Of these, over 70 percent are between family stages in non- adjacent social classes. All of the statistically signifi— . cant differences are from comparisons on stages two, three and four. Sixty percent of the comparisons on Sears and Federal's consumers in family stages in upper middle— lower middle and upper middle—lower classes yield statistically significant differences. In every case 1H7 the lower class consumers prefer Sears and Federal's on locational convenience. This concurs with the results of the social class affect shown earlier in this chapter. It appears that these stores are convenient to lower class shoppers but not to upper middle class shoppers. This is confirmed by examination of the location of Sears and Federal's stores. Comparisons on Hudson‘s result in no statistically significant differences on locational convenience. This may mean that location of Hudson's outlets makes.them convenient to all classes of consumers. CHAPTER VII DEPARTMENT STORE IMAGES: SELECTED DEMOGRAPHIC CHARACTERISTICS AND SHOPPING HABITS This chapter presents additional findings on the affect of demographic characteristics and shopping habits on department store image. It is comprised of three sections. The first shows the affect of race and sex on the image of the test and "ideal" stores. The second presents the affect of attitudes toward shopping and shopping companionship practices on department store image. The third describes the affect of shopping practices on the image of one test store, Hudson's. The Affect of Race and Sex on Department Store Image gage Race is a factor which influences the images con- sumers hold of department stores. The results of the study indicate perceived differences between races, especially with regard to two of the test stores. Table 17 presents the Mann—Whitney U scores and probabilities associated with the null hypotheses that 149 TABLE l7.——Summary of Mann—Whitney U and probabilities associated with null hypotheses between races for image elements of test and "ideal"'stores. White—Non—Whited Department Stores Image Dimensions Hudsons Sears Federal ”Ideal" a e Merchandise 2 —l.2l -0.0A —2.06 -O.7l Suitability pb NSC NS .05 NS Sales 2 -O.45 —2.39 —l.46 —O.86 Personnel p NS .05 NS NS Store 2 —l.88 —2.02e —i.27 —o.u2 Congeniality p NS .05 NS NS e e e — . - . — . 6 - . Locational Z l 10 l 33 2 l l 88 Convenience p NS NS .05 NS aValue of U with n2 > 20. bProbabilities of true null hypotheses. CNot significant at the 5 percent level. dWhite subjects rate store in question higher than non—white subjects, except where noted. eNon-white rate store in question higher than white subjects. 150 there are no perceived differences between races for image dimensions of the test and "ideal” stores. In total, only one—quarter of the comparisons between white and non-white subjects result in differences that are significant at the 5 percent level. But, all of these differences occur on Sears and Federal's where, in each case, statistically significant differences are seen in one—half of the comparisons. The comparisons between races on the individual image dimensions fit the pattern described above. Signifi- cant differences occur in comparisons on Sears and Federal's and indicate that, with the exception of Sears' sales personnel, non—white consumers rate the stores higher than white consumers. White subjects are con— cerned with the quality of sales personnel while non— white subjects emphasize the importance of store con— geniality. Although consumers' perceptions of the competitive strengths and weaknesses of Sears and Federal's differ between races, before management action can be taken to exploit these differences the basis for them must be understood. For example, why do non-white consumers rate Sears' sales personnel low and yet prefer the congeniality of the store? Do they expect poor sales service and so rate the service low regardless of its nature? Are there changes which could be made in the recruiting and training of sales people which would solve this problem? ‘, , .i__-_.—.—___‘, Dami- 151 Federal's appears to be successful in obtaining the non—white consumers preference for merchandise and store location. Should Federal's management investi- gate the white shoppers' merchandise needs so as to better serve these customers? Should Federal's re- evaluate their store location policies to make their new stores more convenient to white shoppers? Comparisons of Hudson's on the four image di— mensions between races reveal no perceived differences. It may be that the diversity of Hudson's stores has re- sulted in satisfying the demands of white and non— white shoppers. It is also possible, however, that the dissimilarity of Hudson's various types of outlets has left the consumers confused about Hudson's so that no preferences are seen in comparisons between races. This question bears investigation. Sex The sex of the subject has only a small influence on consumers' images of department stores. Findings of the research show little evidence of perceived differ— ences in the test and "ideal" stores between sexes. Table 18 presents the Mann—Whitney U scores and probabilities associated with the null hypotheses that there are no perceived differences between sexes for the image dimensions of the test and "ideal" stores. Eighty— two percent of the comparisons between sexes yield 152 TABLE l8.—-Summary of Mann—Whitney U and probabilities associated with the null hypotheses between sexes for image dimensions of test and "ideal" stores. Male-Femaled Department Stores Image Elements Hudsons Sears Federal "Ideal" Merchandise 2a —3.59 —O.l8 —o.27e -O.6l Suitability pb <.01 NS NS NS Sales 2 —O.93 -3.l6 —l.33 —O.54 Personnel p NSC <.01 NS NS Store 2 —l.lS —O.l3 -O.2O ~O.29 Congeniality p NS NS NS NS Locational z —2.12 —o.uue -o.38 —o.19 Convenience p .05 NS NS NS aValue of U with n2 > 20. bProbabilities of true null hypotheses. cNot significant at the 5 percent level. dFemales rate store in question higher than males, \ except where noted. ‘ eMales rate store in question higher than females. 153 differences which are not statistically significant. However, one—half of the comparisons on Hudson's (those on merchandise suitability and locational convenience) result in differences which are statistically significant. It appears that the sex effect is much stronger on Hud— son's than on the other test stores and the "ideal." The direction of the ratings, in comparisons be— tween sexes resulting in significant differences, show female respondents rate Hudson's higher than men on merchandise suitability and locational convenience. Female subjects also rate Sears higher on sales person— nel. Females play a dominant role in consumer purchasing in mid—20th century America.1 Hudson's appeal to the fe— male shOpper may give the store a strong competitive ad— vantage. Sears and Federal's would be wise to evaluate their images to relate more effectively to the female consumer. It might also be profitable for department stores to design areas frequented by men in such a way as to satisfy their special needs. For example, central- izing men's clothing, shoes and accessories would reduce the time and energy now required by the numerous depart- ments scattered throughout the store. lFerdinand F. Mauser, Modern Marketing Management (New York: McGraw Hill Book Company, Inc., 1961), pp. 176—177. 154 The Affect of Shopping Attitudes and Shopping Companionship‘PracticeS‘on Department‘Store'Image Shopping Attitudes Shopping attitudes affect consumer‘s perceptions of particular department stores. The results of the study indicate that subjects who enjoy shopping have images of department stores which are different from those of subjects who do not enjoy shopping. Table 19 presents the Mann—Whitney U scores and probabilities associated with the null hypotheses that there are no perceived differences between consumers grouped by shopping enjoyment for image dimensions of test and "ideal" stores. Two—thirds of the comparisons on the test stores result in differences which are statistically significant. For Hudson's, all of the comparisons between consumers who enjoy shopping and those who do not yield differences which are statistically significant. Three- quarters of the comparisons on Sears result in differences which are significant. However, for Federal's only one comparison, and for the "ideal" none of the comparisons, give statistically significant differences. In all of the comparisons on the test stores re— sulting in significant differences consumers who enjoy shopping score the stores higher than those who do not enJoy shopping. On the ”ideal” store, none of the 155 TABLE l9.——Summary of Mann—Whitney U and probabilitieS' associated with the null hypotheses between consumers classified by shopping enjoyment for image dimensions of test and "ideal"'stores. Enjoy—Don't Enjoyd Department Stores Image Elements Hudsons Sears Federal "Ideal" Merchandise za —3.12 —3.56 -i.6u —o.iie Suitability pb <.01 <.0l NS NS Sales 2 —u.6i -3.54 -3 07 —i.57 Personnel p <.01 <.01 <.01 NS Store 2 —3.20 —l.2O —l.39 -O.15 Congeniality p <.01 NSC NS NS Locational Z ”1'98 ‘2-73 ‘1-77 -O.15. Convenience p .05 <.01 NS NS aValue of U with n2 > 20. bProbabilities of true null hypotheses. CNot significant at the 5 percent level. dConsumers who enjoy shopping rate store in question higher than consumers who don't enjoy shopping, except Where noted. I f eConsumers who don’t enjoy shopping rate store in question higher. 156 comparisons yield significant differences so no indi— cation of the direction of subjects' preferences is available. The shopping attitude affect is stronger for Hudson's and Sears than for Federal's. This may indi— cate that Hudson's and Sears have greater potential for taking actions which encourage those people who enjoy shopping. Activities which tend to encourage customer creativity and provide shopping excitement may prove profitable. Small fashion shows on the floor of various men's and women's clothing departments would bring color and excitement into regular shopping. Home decorating consultants shown furnishing model rooms and explaining the process may make shopping for home furnishings more enjoyable. For those who dislike shopping, department stores may be able to decrease dissatisfaction by making the shopping process as fast and easy as possible. For example, special sales personnel may be assigned to look after the needs of shoppers who are in a hurry. It may be that consumers‘ attitudes towards shopping activities and institutions interact and tend to amplify one another resulting in higher scores for department stores by those who enjoy shopping and vice versa. Shopping Companionship Shopping companionship has little influence on con— sumers' images of department stores. Research findings show few significant differences in comparisons among subjects who like to shop alone, those who like to have company while shopping and those who have no preference. Table 20 presents the summary of Mann—Whitney U scores and probabilities associated with the null hy— potheses that there are no perceived differences among consumers classified by shopping companionship for image dimensions of test and "ideal" stores. Less than 20 per— cent of the comparisons result in differences which are statistically significant. About one—third of the com- parisons between consumers who like to shop alone and those who like to have company when shopping yield statistically significant differences. None of the comparisons between consumers who like to shop alone and those who have no preference on shopping companionship result in significant differences. One—quarter of the comparisons between those who have no preference on shopping companionship and those who like company when shopping yield significant differ— ences. Comparisons on Sears are most affected by a shopping Companionship affect with 50 percent of the comparisons resulting in statistically significant differences. Hudson's and Federal's have only two and one comparisons l58 TABLE 20.-—Summary of Mann—Whitney U and probabilities associated with the null hypotheses between consumers classified as to shopping companion— ship for image dimensions of test and "ideal" stores. Shopping Department Stores C . . c ompanlOHShlp Hudsons Sears Federal "Ideal" Merchandise Suitability Alone— z: —o.56d —2.75d —i.25d -o.08d In Company p NS <.01 NS NS Alone— 2 —o.i5d —i.u6 —0.26 —0.75d Neutral p NS NS NS NS In Company— 2 —0.08 —2.25 —O.82 -O.75d Neutral p NS .05 NS NS Sales Personnel Alone— 2: —i.7ud -2.Oid —0.98d —l.l6d In Company p NS .05 NS NS Alone- 2 —l.6l —l.23 —l.l8 -0.02 Neutral p NS NS NS NS In Company— z —2.30 —l.96 —l.79 —O.29 Neutral p .05 .05 NS NS Store Congeniality , d d Alene— z: —i.6ud —0.79d -2.u7 —2.71 In Company p NS NS .05 .Ol Alone- 2 -O.19 —O.72 —0-75 -0-33 Neutral p NS NS NS NS In Company— 2 -O.SO —l.98 —l.76 —l.l3 Neutral p NS .05 NS NS Locational Convenience d d Alone— 2: .2.05d —2.17d —o.78 —o.05 In Company p .05 .05 NS NS d Alone— 2 —o.6od -i.ou —0.73 -i.32 Neutral p NS NS NS NS d - In Company— z -0.03 ‘1-68 "1‘22 1&36 Neutral p NS NS NS aValue of U with n2 > 20. b Probabilities of true null hypotheses. CConsumers in first group in each comparison rated store in question higher, except where noted. higher. dConsumers in second group in com | 6Not significant at 5 percent level. parison rated store in question 159 respectively which result in significant differences. It may be that shopping at Sears is more conducive to shopping in company than the other test stores. None of the comparisons on "ideal" show a shopping companion— ship affect. There is some indication that those who like to shop in company have more favorable images of department stores. Eighty percent of the comparisons which result in significant differences reveal that those subjects who like to have company while shopping rate the store in question higher. This finding is consistent on all image dimensions except store congeniality where Federal's and "ideal" are rated higher by consumers who like to shop alone. It may be that the higher scores result from social interaction between shopping companions. For example, a person who regularly shops with a group of friends who have a favorable image of a particular de— partment store may tend to adopt that image of the store. Management strategy may be designed to encourage people to shop in groups. Communications and activities stressing the social side of shopping may help to make the shopping activity more enjoyable for more people. This should lead to greater sales assuming that consumers tend to purchase in stores where they find shopping enjoyable. 160 The Affect of Consumer Shopping Practices on the Image of Hudson's Shopping Recency Shopping recency has only a slight influence on the image consumers have of Hudson's. The results of the study show little evidence of perceived differences of Hudson's among subjects grouped by recency of shopping there. Table 21 presents the summary of the Mann—Whitney U scores and probabilities associated with the null hy— potheses that there are no perceived differences among consumers classified by shopping recency at Hudson's for image dimensions of Hudson's. Only 25 percent of the comparisons between groups of consumers classified by shopping recency yield differences which are statistically significant. None of the comparisons of the images be- tween consumers in group one (most recent) with those in group two (intermediate reCency) resulted in statistically significant differences. However, 50 percent of the com— parisons between groups one and three (least recent) and 25 percent of the comparisons between groups two and three give differences which are statistically signifi— cant. The direction of the ratings, where comparisons result in significant differences, indicate that the shorter the period between shopping and questioning the ,5: y 161 so mmanfl mCOmosm popes QOmHstEoo Ho COHnoQ mEHp peoomm CH maoESmcoo .Hm>mH pamosmg m can on pQMOHMHcmHm poZo .QOHQOEHU mmmEH p .popoq ohms: pmmoxm aCOHpmmSU CH QOHonEHU mmmEH so nmgmHg msomodm popes zomHstEoo homo mo oOHLoQ oEHp pmsHa CH mnmESmcooo .womoflpommg HHSC mosh wo moHpHHHhmposmp .mm A C QpHS D mo mSHd>w mo. mz mz mz Q owe npcoa HA mo.m1 mw.H| mm.ol ms.HI N Iowa gpcos HA use sows HA Ho.v m2 m2 me. e owe asses HA ms.mu No.0: mo.ou sm.mn N Iowa some Hv m2 m2 m2 mmz pg own Spfloa Hv use xmos HA mm.HI Umm.o| 602.01 Hw.o| MN Iowa xoos Hv ooQoHcm>Qoo thHmHsomsoo Hoccomsmm szHHoprsm he: mooa moo who mm m o comm H .p a pm H m omHocmgosmz wcHooogm mQOHmcoEHQ mwmaH llllmwwwmmflllll .oQOBm pmmp m hog mCOHonsHU mmmEH now mosmomn quamogm on we coHMHmmmHo mamazmqoo memento mommgpomag HHBQ one ngs ompmHoommm moHpHHHpmoonQ new 3 mocpHQZIQQwE mo msmEE5m11.Hm mqmoH pcoomog m on» no acmOHMHcmHm poZU .GOprmdv CH QOHonEHU mmmEH QOM msomm osooow gmnp pongg mcomosm mums comHsmasoo Some mo 930nm pmus CH mnoEquooo .momospogmg HHSQ wasp wo moHpHHHQMQoan .om A Ne gene 2 no esHe>e mo. Ho.v mz Ho.v . a heetz om.ml :m.m| wm.0| mm.m1 N IHmQOHmmooo Ho.v H0.v mz Ho.v a no>mz :N.ml mm.m: 0m.0| Hs.m| N IHmzoq H0.V .mo. omz H0.v no HmQOHmmooo mm.:| Nm.ml no.0: mo.m| mN IHmmOH moeoHco>soo mpHHmHemwcoo qucomnmm szHHnprom ApHmmoA «cos moo who mm m H .p H pm H m mmHUcmgosoz owCHancm mQOHmcoEHQ omeH msomosm .wQOHmemEHU omeH pom os0pm pmop w on szszH mQHQmogm on we ooHMHmmMHo mnoESmsoo comapmo mmmogcogza HHse was spas empesoomme merHHHheeote ese o seepaeziszez so assessmln.mm mamas \ 16A comparisons among consumers grouped by their shopping loyalty to Hudson's, differences are found which are statistically significant. Only those comparisons on the sales personnel dimension yield differences which are not statistically significant. In every case where significant differences occur the first (most loyal) group in each comparison rate Hudson's higher on the image dimension in question. It was anticipated that loyal customers would hold more favorable images of a department store than other consumers. Loyal shoppers should be more aware of the various aspects of the store and may be psychologically com— mitted to the store. Martineau describes a "halo effect" of shopping loyalty which causes the consumer to forget or overlook any undesirable qualities of the store.2 The differences in the images held by loyal, occasional and non—shoppers may be useful to the store manager in developing his marketing plans. Promotional and other merchandising activities aimed at the loyal customer should concentrate on the qualities of the pro- ducts and services available. This will give the loyal shopper support for purchases already made and encourage him to buy again. Promotion to the non-loyal consumer should be designed to sell the store as a good place to 2Martineau, ”Sharper Focussfor the Corporate Image," LO ' C_i_t'> p0 53 165 shop. This will encourage the non—loyal shopper to sample what the store has to offer. »The findings of this study suggest the value of further research on the development, maintenance and decline of customer loyalty. How do consumers' atti- ‘ tudes toward stores change? Is it possible to describe in demographic or psychological terms persons who are more or less likely to be loyal? Do shoppers classified by loyalty have distinct shopping patterns within the store? Answers to these and other questions would assist in the understanding of consumer images. Cash versus Credit Consumers' perceptions of Hudson's are only slightly influenced by whether they pay for their purchases by cash or charge. Study results show little evidence of differentiable images of Hudson's among subjects grouped by payment method. Table 23 shows the summary of the Mann—Whitney U scores and probabilities associated with the null hy— potheses that there are no perceived differences between consumers classified by their method of payment to Hudson's for image dimensions of Hudson's. Only one of the four comparisons between cash and Charge customers results in a difference which is Statistically significant. In this comparison the charge 166 .Hm>mH scooped m one an pcmOHMHcmHm uOZo .coapmmsv cH QOHmcmEHo mm mmmEH CO mhmEOPmfio mwMMQO GQQP thMHQ mCOmUSE Umpmk mhmfiopmfio S DU .COHpmmsv CH QOHmcmEHo wmmEH so mLoEOUmSo ammo swap nmngn whompsm poems mmeoumSO owpmnoo .mmmogpogmg HHSQ comp to mmeHHHQmposmh .om A mg QpHa 0 mo osHo>m mz mz emz me. he manage mm.o| mm.H- eem.ou Hm.mu eN lemme ooQoHso>Qoo mpHHchomcoo Hoccomsmm mpHHHpmpHom vogue: HMQOHpmooH onOpm monm mmHochoamz momEmmm mCOHmcmEHQ mmeH mQOmosm .mQOHQOEHo mmmeH com msOpm who» m on pcmEmwg mo tempos nHogp on we ooHaHmmmHo msoESmcoo Semapop mommgpommg HHse the ass: empeHQOmme meHeHHaeeeoha see b amessez-eeez so sttEEgmuu.mm mamae 167 customers rate Hudson's higher than the cash customers on merchandise suitability. It was assumed that charge account customers would be more closely associated with the store and so hold different images than cash customers. It was also assumed that charge account customers would be more loyal than cash customers. However, the results show that loyal customers have favorably different images of Hudson's while charge account customers do not. These findings cast doubt on the use of charge accounts to build customer loyalty. Shopping Location Shopping location within Hudson's does not affect consumers' perceptions of the store. The research find— ings indicate few significant differences in comparisons of consumers grouped by their shopping location within Hudson's. Table 2A shows the summary of Mann—Whitney U scores and probabilities associated with the null hypotheses that there are no perceived differences among consumers classified by shopping location within Hudson's for image dimensions of Hudson's. Less than 20 percent of the comparisons result in statistically significant differ— ences. The two comparisons which do yield significant differences are on sales personnel. Both of these comparisons are with people who shop exclusively upstairs 168 C0 CmCmHC mCompsm opmm ComHmeEoo Mo asoam UCoom .Ho>oH pCmoCmQ m on pm qu C .COprmCU CH COHowEH MCp CoCmHC mCowpsm ome COmHCmCEoo Comm mo Coonm pm .mmmmeong HHCC comp mo moH .COHpmoCU CH C m CH mCmESmCooo oHCHCmHm poZm OHmCoEHp mmmEH p omeH Co goosm UCOomm CHC CH mCoECmCooo eHHHheeosmh .om w NC CpHs 0 Com ous>w lIllllllllIIiiIlillllIllllill mz wz Q wCHmmeSOQ mz wz oCm mCHNmeD mz wz H0.v mz Q mCHmmezoa oCm mCHmpmgb emm.Hu emH.ou emm.mu ewm.ou N -mCHehma: m2 m2 H0.v 02 a mCprmCaom omz.o: OHH.0| 0mm.m| 003.0: N nmCHmmeD ooCoHCo>Coo szHmHCmMCoo HoCCOmCmm mpHHHQmpHsm mCOHmCmEHQ odeH COHpmooq wCHQmoCm mCOmosm .mCOHmCoEHU owwaH Com mCOQm pmop a CHCsz COHBCOOH wCHQQOCm on we UmHCHmmmHo mCoECmCoo Coozpoh momonoghC HHSC the CeHz empeHeomme merHHHQeeona est : aeCpngnccmz no assessmuu.em mamae 169 and both indicate that the upstairs shopper rates Hudson's lower on sales personnel. Customers who shop exclusively upstairs apparently are more exacting in their needs for personal service in Hudson's than other customers. It may be that the type and quality of merchandise they purchase requires special sales personnel assistance. For example, departments which sell luxury items such as jewelry, high quality furniture and furs require special sales service. It is also possible that the self images of these consumers cause them to feel that they deserve better service. Analysis of the customer group which shops exclusively Upstairs at Hudson's to discover their demographic and shopping characteristics may be valuable. Advertising Readership Consumers' advertising reading practices influence their perception of Hudson's. The results of the study show significant differences in comparisons among con- sumers grouped by readership of Hudson's advertisements. Table 25 presents the Mann—Whitney U scores and '. probabilities associated with the null hypotheses that there are no perceived differences among consumers Classified by their readership of Hudson's advertisements for image dimensions of Hudson's. Almost 60 percent of the comparisons result in differences which are 170 .Ho>oH pCooCoQ m pm pCMoHQHCme p020 .COprmCU CH COHwCoEHp mmmEH Co QCOCm UCoomw Cme CmCmHC mComosm mums ComHCmQEoo Como Co QSOCw meHm on CH mCoECmCooo .mmmmCCOQzC HHSC comp mo mmeHHHCMCOCmC .om A me est a Co msHe>e wz m2 m2 pmz Q Cm>mz mm.0| mm.Hl 0H.H| 00.0: N ImoEHpoEow mz H0.v m0. m0. Q Co>mz mN.Hu Hw.m- :m.m- em.ml N -mgezaa m0” H0.v H0.V H0.v hQ mmEHpoEom mm.ml 0H.:I wm.m| :0.:| wN Immm3H< moCoHCm>Coo szHmHCmmCoo HoCComCom mpHHHompHdm mpCmEomeCo>©< HmCOHpmooH oCopm monm meUCmCoCoz o . m.ComUCm mo QHCmCmomom mCOHmCoEHQ owmsH mCompsm .mCOHmCoEHU ommEH Com mCHmeCo>om m.oC0pm pwmp Co QHCmCopmoC op mm ponHmmeo mCoECmCoo Comspoo mmmonOQzC HHse the CpHa etpmHeowma meeHHHeenosa ese o sthHezuCCez Co Chasesmus.mm names 171 statistically significant. Seven of the eight comparisons of consumers who always read Hudson's advertisements with those who sometimes do and those who never do, yield statistically significant differences. All of the com- parisons which result in significant differences indi- cate that the first group in each comparison (the more avid reader) rates Hudson's higher on each image di— mension. The results clearly point out that those con— sumers who read Hudson's advertisements perceive the test store on the image dimensions as different than those who tend not to read the advertisements. The advertising readership affect on consumers' images of Hudson's may result from selective exposure, perception and memory of Hudson‘s advertisements. This might be explained in terms of Festinger's theory of 3 in that those who prefer Hudson's cognitive dissonance tend to seek support for their views. The results of the advertising readership affect coincide with those of Arons which show viewers of a department store's advertising A have more favorable images than non-viewers. 1 Avid readers of Hudson's advertisements may be loyal Hudson's customers and non-readers may be non— customers. This should be investigated. If the assum— ption is correct it lends support to the recommendation 3Festinger, op. cit. “Arons, op. cit., p. ll. made above that promotion to customers and non—customers should be based on different objectives. Social Support Social support for consumers' beliefs about Hudson's have an affect on their perception of that store. Results of the research show significant differences in comparisons between consumers who have social support for their be— liefs and those who do not. Table 26 shows the summary of the Mann—Whitney U scores and probabilities associated with the null hy— potheses that there are no perceived differences between consumers classified by the social support for their be— liefs about Hudson's for image dimensions of Hudson's. Half the comparisons result in statistically significant differences. Comparisons on store congeniality and locational convenience give differences which are statis— tically significant. Comparisons on merchandise suita— bility and sales personnel do not yield significant differ— ences. All comparisons yielding significant differences indicate that those who have social support (friends agree) for their beliefs about Hudson's rate the store higher than those without social support. Although the results are somewhat inconclusive they do tend to support the notion that individuals who have social support for their feelings about a department store will have more definite images of that store. 173 .Ho>oH pCmonQ m on pm pCmoHQHCme p020 .COHpmoCU CH COHmCmEHU mmmEH Co CmaHCerompsm mums m.Comp5m poops mmCHHoom CHon CpHs omsmm mpCmHCC omOCs mCoECwCooo .mmmoCCOQzC HHCC mCCp mo mmeHHHCmnonmh .om A Ne gene a no esHe>e Ho.v Ho.v m2 emz he mesmewHo meetHCm Hm.m| ow.m| 0:.Hu :H.HI . mN Immam< mUCmHCQ eeCeHeeeeoo ngHeHCeweoo HeCCOmCem spHHHeepHsm phoaesm HmCOprooq maoum monm omHoCmCosz onHoom wCOHmCoEHQ mmmEH mCompsm .mCOHmCmEHp mmmEH Com oCOpm pwou a cache mCoHHon CHon Com pCOQQSm HmHoom on mm pmHmHmmmHo mCmECmCoo Coozpop mommCCOQmC HHCC Csz ompwHoommm moHpHHHomCOCQ pr D zonHCBICCwS Co hCmEECmII.mm mHm. 159—169. Edwards, Allen L. "The Relationship Between the Judged Desirability of a Trait and the Probability That the Trait Will Be Endorsed," Journal of Applied Psychology, XXXVII (1953), 90—93. Edwards, A. L., and K. P. Kilpatrick. "A Technique for the Construction of Attitude Scales," Journal of Applied Psychology, XXXII (19A8), 37A—3 . Evans, Franklin B. "The Brand Image Myth," Business Horizons, IV (Fall, 1961), 26. Ferber, Robert. "Our Changing Consumer Market," Business Horizons, I (Spring, 1958), A9—66. Ferber, Robert. "Research on Household Behavior," American Economic Review, LII (March, 1962), 19—63. Fisk, George. ”A Conceptual Model for Studying Customer Image," Journal of Retailing, XXXVII (Winter, 1961- W 3 Frank, Laurence K. "Suggestion for A Theory of Learning," Psychological Review, XXX (1923), 1A5—1A8. Frank, Laurence K. ”The Problem of Learning,” Psychological Review, XXXIII (1926), 329—351. Fullerton, G. S. "In What Sense Two Persons Perceive the Same Thing,” Philosophical Review, XVI (1907), 506-518. 203 Gardner, Burleigh B., and Sidney J. Levy. "The Product and the Brand," Harvard Business ReView, XXXIII (March—April, 1955), 33—39. Gardner, Burleigh B., and Lee Rainwater. "The Mass Image of Big Business," Harvard Business Review, XXXIII (November—December, 1955), 61—66. Glick, Paul C. ”The Family Life Cycle," American Sociological Review, XII (April, 19A75, 16A—17A. Gordon, Kate. "Perception and Imagination,” Psychological Review, XLI (1935), 166—185. Greenberg, Allan. "Frame of Reference of Image ReSponses," Journal of Marketing, XXV (April, 1961), 62-6A. Gray Matter. (Retail Edition). Grey Advertising, Inc., II (September, 1965). Guttman, L. ”The Third Component of Scalable Attitudes," International Journal of Opinion and Attitude Research, IV (1950), 285—287. Haggerty, M. E. "The Laws of Learning," Psychological Review, XX (1913), All—A22. Heidingsfield, Myron S. "Why Do People Shop in Downtown Department Stores," Journal of Marketing, XIII (April, 19A9), 510—512. Hicks, G. Dawes. "On the Nature of Images," British Journal of Psychology, XV (October, 192A), 121—1A8. Hill, Edward W. "Corporate Images Are Not Stereotypes," Journal of Marketing, XXVI (January, 1962), 72—75. Holt, Henry R. "Imagery: The Return of the Ostracized," American Psychologist, XIX (March, 196A), 25A—26A. Hovland, C. I., and M. Sherif. "Judgmental Phenomena and Scales of Attitude Measurement: Item DiSplace— ment in Thurstone Scales," Journal of Abnormal and Social Psychology, XLVII (1952), 822—832. Kahl, Joseph A., and James A. Davis. "A Comparison of Indexes of Socio—Economic Status," American Sociological Review, XX (June, 19555, 313—325. Kantor, J. R° ”Suggestions Toward a Scientific Inter— pretation of Perception," Psychological Review, XXVII (1920), 191—216. 20A Kluver, Heinrich. "An Experimental Study of the Eioetic Type," Genetic Psychology Monograph, March, 1926. Laird, Donald A. "How the Consumer Estimates Quality By Subconscious Sensory Impressions," Journal of Applied Psychology, XVI (1932), 2A1—2A6. Lansing, John B., and Leslie Kish. "Family Life Cycle As An Independent Variable," American Sociological Review, XXII (October, 1957), pp. 512—519. Lapiere, R. T. "Attitudes Vs. Actions," Social Focus, XIV (193A), 230—237. “"“‘ Likert, R. "A Technique for the Measurement of Atti- tudes," Archives of Psychology, #1AO, 1932. Martineau, Pierre. ”It's Time to Research the Consumer,” Harvard Business Review, XXXIII (July—August, 1955),‘ A5-5A. Martineau, Pierre. "The Personality of the Retail Store," Harvard Business Review, XXXVI (January—February, 1958),137-55. Martineau, Pierre. "Sharper Focus for the Corporate Image,” Harvard Business Review, XXXVI (Nobember— December, 1958), A9—58. Martineau, Pierre. ”Social Class and Spending Behavior," Journal of Marketing, XXIII (October, 1958), 121-130. Maslow, A. H. ”A Theory of Human Motivation," Psycho- logical Review, L (July, 19A3), 370-396- McDermott, L. M. "Why People Buy At Department Stores," Journal of Marketing, L (July, 1936), pp. 53—55. Meyer, Harold D. "The Adult Cycle," The Annals of the American Academy of Political and Social Science, CXIII (September, 1957), 58-67. Mindak, William A. ”A New Technique for Measuring Advertising Effectiveness," Journal of Marketing, XX (April. 1956), 367—378. Mindak, William A. "Fitting the Semantic Differential to Marketing Problems," Journal of Marketing, XXV (April, 1961), 28-33. 205 Munn, Henry L. "Brand Perception as Related to Age, Income and Education," Journal of Marketing, XXIV (January, 1960), 29-3A. Myers, Garry C. "A Comparative Study of Recognition and Recall," Psychological Review, XXI (191A), Myers, Robert H. "Sharpening Your Store Image," Journal of Retailing, XXXVI (Fall, 1960), 129—137. Nelson, Bardin H. "Seven Principles in Image Formation," Journal of Marketing, XXVI (January, 1962), 67—71. Nixon, M. K. "Notes On the Measurement of Consumer Attitudes," Journal of Marketing, I (July, 1936), pp. 13‘19o Odiorne, G. S. "A Search for Objectives in Business—- The Great Image Hunt," Michigan Business Review, XVIII (January, 1966), 19—2u. Ogden, Robert Morris. ”Imageless Thought Resumé and Critique," Psychological Bulletin, VIII (June 15, 1911), 183—197. Osgood, C. E., and G. J. Susi. "Factual Analysis of Meaning," Journal of Experimental Psychology, L (1955), 325—338. Paranka, Stephen. "Marketing Predictions from Consumer Attitudinal Data," Journal of Marketing, XXV (July, 1960), A6—51. Park, R. E. "Experience and Race Relations," Journal of Applied Sociology, IX (192A), 21—22. Peterson, Joseph. ”The Functioning of Ideas in Social Groups," Psychological Review, XXV (1918), 21A—226. Pirie, Margaret C. ”Marketing and Social Classes: An Anthropoligists View,” The Management Review, XLIX (September, 1960), pp. A5—A8. Reiser, Oliver L. "The Structure of Thought," Psycho- logical Review, XXXI (January, 192A), 51-73. Rich, Stuart U., and Bernard Portis. "Clues for Action from Shopper Preferences," Harvard Business Review, XLI (March—April, 1963), l32—1A9. 206 Rich, Stuart U., and Bernard Portis. "The Imageries of Department Stores," Journal of Marketing, XXVIII (April, 196A), 10—15. Rogers, A. Sophie. "Auditory and Tactual Perceptions: The Role of Image,” American Journal of Psychology, XXXIV (April, 1923), 250—266. Rosander, A. C. "An Attitude Scale Based Upon Behavior Situations," Journal of Social PSychology, VIII (1937), 3—15. Roseborough, Howard. "Sociological Dimensions of Con— sumer Spending," Canadian Journal of'ECOnomICS and Political Science, XXVI (August, 1960), pp. A52—A6A. Rosenow, Curt. "The Genise's of the Image," Psychological Review, XXV (1918), 297—3OA. Russell, Bent S. "Brain Mechanisms and Mental Images," Psychological Review, XXVII (1920), 23A—2A5. Schatzman, Leonard, and Anselm Strauss. "Social Class and Modes of Communication," The American Journal of Sociology, LX (January, 1955), 329—338. Secor, W. B. "Visual Reading: A Study of Mental Imagery," American Journal of Psychology, XI (January, 1900), 225—236. Sherif, Muzafer. "A Study of Some Social Factors in Perception," Archives of Psychology, July, 1935, 5—60. Sherif, Muzafer, and Hadley Contril. "Psychology of Attitudes, Part II," Psychological Review, LIII (January, 19A6), 1—2A. Sherif, Muzafer, and C. I. Hoveland. "Judgmental Phenomena and Scales of Attitude Measurement: Placement of Items with Individual Choice of Number of Cate— gories," Journal of Abnormal and Social Psychology, XLVIII (1953), 135—1A1. Smith, W. G. "The Place of Repetition in Memory," Psychological Review, III (1896), 21—31. Spector, Aaron J. "Basic Dimensions of the Corporate image," Journal of Marketing, XXV (October, 1961), 7—51. 207 Stanley, Hiram M. "Language and Image,"'Psycnological Review, IV (1897), 67—71. Straits, Bruce C. "The Pursuit of the Dissonent Con— sumer," Journal of Marketing, XXVIII (July, 196A), 62—66. Strong, Edward K., Jr. "The Effect of Length of Series Upon Recognition Memory," Psychological Review, XIX (1912), AA7—A62. Strong, E. K., Jr. "The Affect of Time Interval Upon Recognition Memory," PsychologiCal Review, XX (1913), 339—371- Strong, E. K., Jr. "The Effect of Size of Advertisements and Frequency of Their Presentation," PsychologiCal Review, XXI (191A), 136—152. ‘““‘ Thurstone, L. L. "Theory of Attitude Measurement," Psychological Review, XXXVI (1929), 222—2A1. Toch, Hans, and Allen Tensure, and William Lazer. "The Public Image of Cancer Etiology," Public Opinion Quarterly, XXV (1961), All—AlA. Tolman, Edward Chace. "Purpose and Cognition: The Determiners of Animal Learning," Psychological Review, XXXII (1925), 285—297. Tucker, W. T. ”How Much of the Corporate Image Is Stereotype"? Journal of Marketing, XXV (January, 1961), 61—65. Tyler, William D. ”The Image, The Brand, The Consumer," Journal of Marketing, XXII (January, 1958), 162—165. Vernon, Raymond. ”Predetermining the Necessary Size of a Sample in Marketing Studies," Journal of Market— ing, II (July, 1937), pp. 9-12. Walker, Kenneth F. "The Nature and Explanation of Behavior,” Psychological Review, XLIX (19A2), 569—585. Weale, W. Bruce. "Measuring the Customer's Image of a Department Store," Journal of Retailing, XXXVII (Summer, 1961), A0—A8. Wells, William D., F. G. Goi, and Stuart Seader. "A Change In Product Image," Journal of Applied Psychology, XLII (1958), 120—121. 208 Wiebe, G. D. "Some Implications of Separating Opinions from Attitudes," Public Opinion Quarterly, XVII (Fall, 1953), 328—336. Winch, W. H. "The Function of Images," Journal of Philosophy, Psychology and Scientific Methods, V (1908), 337—352. Wolgast, Elizabeth H. "Do Husbands or Wives Make the Purchasing Decisions"? Journal of'Marketing, XXIII (October, 1958), 151-158. Essays and Articles in Collections Allport, G. W. "Attitudes," A Handbook of Sooial PsychOIOgy, Carl Murchison, editor. Worchester, Massachusetts: Clark University Press, 1935. Pp. 798—8AA. Allport, G. W. "The Historical Background of Workers Social Psychology," Handbook of Social Psychology, Vol. 1, Theory and Method, G. Lindzey, editor. Cambridge, Massachusetts: Addison-Wesley, 195A. Pp. 36—56. Barton, S. G. "The Life Cycle and Buying Patterns," Consumer Behavior, Vol. 1;, Lincoln Clark, editor. New York: New York University Press, 1955), Pp. 53-57. Beldo, Leslie A. "What Consumer Profiles Don't Tell Us About Who's Buying Your Product," Advancing Market— ing Efficiency, Lynn H. Stockman, editor. Chicago: American Marketing Association, 1959. Pp. 128—1A7. Beldo, Leslie A. "Introduction to Attitude Research and Management Decisions," Effective Marketing Coordi- nation, George L. Baker, Jr., editor. Chicago: Amirican Marketing Association, 1961. Pp. 583— 59 . Bernard, L. L. "Attitudes, Social," Encyclopedia of the Social Sciences, E. R. A. Seligman and Alvin John- son, editors. New York: The Macmillan Company, 1930. Pp. 305-306. Bliss, Perry. ”Non Price Competition at the Department Store Level," Marketing in Transition, Alfred L. Seelye, editor. New York: Harper and Brothers, 1958. Pp. 161—170. 209 Bourne, Francis S. "Different Kinds of Decisions and Reference—Group Influence," Marketing and the Behavioral Sciences, Perry Bliss, editor. Boston: Allyn and Bacon, 1963. Pp. 2A7—255. Brown, George H. "Brand Images Among Low Priced Cars," Marketing Concepts and Changing Times, Richard M. Hill, editor. Chicago: American Marketing Association, 1960. Pp. 61-65. Brown, Lyndon O. "Principle Psychological Techniques in Motivation Research,” Motivation and Market Behavior, Robert Ferber and Hugh G. Wales, editors. Homewood, Illinois: Richard D. Irwin, Inc., 1958. Pp. 76—88. Cannel, Charles. "A Psychologist's View," Motivation and Market Behavior, Robert Ferber and Hugh G. Wales, editors. Homewood, Illinois: Richard D. Irwin, Inc., 1958. Pp. 3—11. Cantril, Hadley. "The Nature of Social Perception," Human Behavior from a Transactional Point of View, F. P. Kilpatrick, editor. Hanover, New Hampshire: Institute for Associated Research, 1953. Pp. 223—232. Carlson, Robert O. "The Nature of Corporate Images," The Corporation and Its Publics, John W. Riley, Jr., editor. New York: John Wiley and Sons, Inc., 1963. Pp. 2A-A7. Child, Irvin L. "Socialization," Handbook of Social Psychology, Vol. 11, Gardner Lindzey, editor. New York: Addison—Wesley, 195A. Pp. 655—692. Coleman, Richard P. "The Significance of Social Stratification in Selling," Marketing and the Behavioral Sciences, Perry Bliss, editor. Boston: Allyn and Bacon, Inc., 1963. Pp. 156—171. Ferber, Robert. "Projective Techniques from an Analyti— cal Point of View," Motivation and Market Behavior, Robert Ferber and Hugh G. Wales, editors. Home— wood, Illinois: Richard D. Irwin, Inc., 1958. Pp. 133—1A3. Fisher, Janet A. ”Family Life Cycle Analysis in Research on Consumer Behavior," Consumer Behavior, Vol. II, Lincoln Clark, editor. New York: New York Uni— versity Press, 1955. Pp. 28—35. 210 Gardner, Burleigh B. "Qualitation Research and Brand Image," Marketing Concepts in Changing Times, R. M. Hill, editor. Chicago: American Marketing Association, 1960. Pp. 5A—59. Gardner, Burleigh B. "Behavioral Sciences as Related to Image Building," New Directions in Marketing, F. S. Webster, editor. Chicago: American Market- ing Association, June, 1965. Pp. 1A5—150. ' Green, Bert F. "Attitude Measurement," Handbook of Social Psychology, Gardner Lindzey, editor. Cambridge, Massachusetts: Addison-Wesley Pub— lishing Company, Inc., 195A. Pp. 335—365. Grubb, Edward L. "Consumer Perception of Self Concept and Its Relation to Broad Choice of Selected Pro— duct Types," Marketing and Economic Development, Peter D. Bennett, editor. Chicago: American Marketing Association, 1965. Pp. A19—A22. Haley, Russell 1. "New Insights Into Attitude Measure— ment," New Directions in Marketing, F. S. Webster, editor. Chicago: American Marketing Association, 1965. Pp. 309—330. Hastorf, A. H. and A. L. Knutson. "The Nature of Attitude and Opinion," Human Nature from the Transactional Point of View, F. P. Kilpatrick, editor. Hanover, New Hampshire: Institute for Associated Research, 1953. Pp. 233—235. Heidingsfield, M. S. "Building the Image-~An Essential Marketing Stratagem," New Directions in Marketing, F. W. Webster, editor. Chicago: American Market— ing Association, June, 1965. Pp. 133-1AA. Herzog, Herta. "Behavioral Science Concepts for Analyz— ing the Consumer," Proceedings of the Conference of Marketing Teachers from Far Western States, Delbert J. Duncan, editor. Berkeley: University of California, 1958. ' KPUBman, Herbert E., and Eugene L. Hartley. "The Learn— ing of Tastes," Marketing and the Behavioral Sciences, Perry Bliss, editor. Boston: Allyn and Bacon, Inc., 1963. Pp. 102—11A. Lansing, J. B., and J. N. Morgan. "Consumer Finances Over the Life Cycle," Consumer Behavior, Vol. II, Lincoln Clark, editor. New York: New York Uni— versity Press, 1955. Pp. 36-51- 211 Levy, Sydney J. "Symbols By Which We Buy," Advancing Marketing Efficiency, Lynn Stackman, editor. Chicago: American Marketing Association, 1958. Pp. AO9—A16. Lucus, Darrell B. "Can the Clinical Techniques Be Validated"? Motivation and Market Behavior, Robert Ferber and Hugh G. Wales, editors. Homewood, Illinois: Richard D. Irwin, Inc., 1958. Pp. . 122—132. Martineau, Pierre. "The Changing American Consumer,” Marketing In Action: Readings, William J. Shultz and Edward M. Mazze, editors. Belmont, California: Wadsworth Publishing Company, 1963, Pp. 30-36. Mayer, Kurt. "Diminishing Class Differentials in the United States," Marketing and the Behavioral Sciences, Perry Bliss, editor. Boston: Allyn and Bacon, 1963. Pp. 185—207. Miller, Donald L. "The Life Cycle and the Impact of Advertising,” Consumer Behavior, Vol. II, Lincoln Clark, editor. New York: New York University Press, 1955. Pp. 61—65. Morgan, James. "A Review of Recent Research On Consumer Behavior," Consumer Behavior, Vol. III, Lincoln Clark, editor. New York: Harper and Brothers, Publishers, 1958. Pp. 93-219. Murphy, Gardner. ”Social Motivation," Handbook of Social Psychology, Gardner Lindzey, editor. Cambridge, Massachusetts: Addison—Wesley Pub— lishing Company, Inc., 195A. Pp. 601—619. Riley, John W., Jr. "The Nature of the Problem," The Corporation and Its Publics, John w. Riley, JET, editor. New York: John Wiley and Sons, Inc., 1963. Pp. 1—11. Rose, Alvin W. "Motivation Research and Subliminal Advertising," Marketing and the Behavioral Sciences, Perry Bliss, editor. Boston: Allyn and Bacon, 1963. Pp. Ago—503. Scheerer, Martin. "Cognitive Theory," Handbook of Social Psychology, Gardner Lindzey, editor. Cambridge, Massachusetts: Addison-Wesley Pub— lishing Company, Inc., 195A. Pp. 97—119. 212 Sherif, Muzafer. "A Preliminary Experimental Study of >Intergroup Relations," Social Psychology At the Crossroads, John H. Rohrer and Muzafer Sherif, editors. New York: Harper and Brothers Pub- lishers, 1951. P. 408. Shibutani, Tomotsu. "Reference Groups as Perspectives," Marketing and the Behavioral Sciences, Perry Bliss, editor. Boston: Allyn and Bacon, 1963. Pp. 220- 23M. Weschler, Irving R., and Raymond E. Bernberg. "Pro- jective Techniques in Attitude Measurement," Motivation and Market Behavior, Robert Ferber and Hugh G. Wales, editors. Homewood, Illinois: Richard D. Irwin, Inc., 1958. Pp. l03-l22. Young, Erle Fisk. "Balance and Inbalance in Personality," Social Attitudes, Kimball Young, editor. New York: Henry Holt and Company, 1931. Pp. 78-86. Unpublished Materials Atkin, Kenward Louis. "Communications Patterns and Effect in Super Market Choice." Ph.D. Thesis, Michigan State University, 1961. Carter, Robert Newton. "The Corporate Image As It Reflects Firm Self—Conception and Affects Patronage Motives." Ph.D. Thesis, University of Florida, 1965. Loewer, Robert A. "A Study of Consumer Perception of Department Stores and Department Store Prices." Ph.D. Thesis, Stanford University, 1965. Rokeach, Milton. "Attitude, Attitude Change, and Behavior Change." 1966. (Prepublication mimeograph copy.) , "The Store Image——A Philadelphia View." Strawbridge and Clothier, A. M. C. Store Principals Meeting, Greenbrier, October, 1959. (mimeographed.) APPENDICES 213 APPENDIX A HOUSEHOLD SIZE 21“ 9.4.1 215 Household Size* Household Size Number of Households Per Cent 1 or 2 members 381,550 33.0 3 or 4 members 428,950 37.1 5 or more members 345,700 29.9 Total 1,156,200 100.0 *Second Biennial Survey of the Metropolitan Detroit NeWSpaper Audience (The Detroit News: Detroit, 1966), p. 12. APPENDIX B PRESENCE OF CHILDREN 216 ‘14 217 Presence of Children* Presence of Children Number of Children Per Cent Youngest child under 6 360,730 31.2 Youngest child 6 — 11 172,270 14.9 Youngest child 12—17 137,590 11.9 Any child under 18 420,460 58.0 No child under 18 291,850 42.0 *Second Biennial Survey of the Metropolitan Detroit News a er Audience (The Detroit'News: Detroit, 19625, p. 12. APPENDIX C HOUSEHOLD INCOME 218 Household Income* Household Income Number of Households Per Cent Under $3,000 $ 3 — 4,999 $ 5 - 7,999 $ 8 - 9,999 $10 — 14,999 $15 - 24,999 $25,000 or more 112,150 115,620 322,580 200,020 240,490 83,250 30,060 9.7 10.0 27.9 17.3 20.8 7.2 2.6** *Second Biennial Survey of the Metropolitan Detroit Newspaper Audience (The Detr01t News: Detroit, 1966), p. 12. **0utside maximum sampling error tolerance. APPENDIX D HOUSEHOLD TENURE, VALUE, MONTHLY RENT 220 221 Household Tenure, Value, Monthly Rent* Owner Occupied ' Total Number Home Value of Households Per Cent Under $7,500 58,970 5.1 $ 7,500 — 9,999 107,530 9.3 $10,000 - 12,499 173,430 15.0 $12,500 — 14,999 164,180 14.2 $15,000 — 19,999 220,830 19.1 $20,000 — 24,999 77,470 6.7 , $25,000 or more 89,030 7.7 Renter Occupied Monthly Rent ggfigzioigs Per Cent Under $50 28,910 2-5** A $ 50 _ 74 134,120 11.6 $ 75 _ 99 62,440 5.4 $100 - 149 25,440 2.2** $150 or more 15,030 1°3** *Second Biennial Survey of the Metropolitan Detroit Newspaper Audience (The Detroit News: Detroit, 1966), p. 1 . **Outside maximum sampling error tolerance. APPENDIX E QUESTIONNAIRE 222 MICHIGAN STATE UNIVERSITY RESEARCH Michigan State University East Lansing, Michigan Consumer Shopping Habits in the Detroit Area Hello, I'm ______ from Michigan State University Research. We are conducting a survey of shopping habits in your area. I would like ymlto help us by answering some questions. lJ” Does any member of your immediate family work for a department store in the Detroit area? .1 5 Yes .2. 17 No. .3 J D/K, N/A If N2 continue interview. If Egg discOntinue inter— view. Explain the purpose of the study and thank the individual for his assistance. This is an in— complete interview. Interviewer should consult instructions for locating additional subjects. INSTRUCTIONS FOR ANSWERING THE SEMANTIC DIFFERENTIAL The purpose of this study is to measure the meanings of certain You will be asked to judge these things In giving your answers, f what these things things to various people. against a series of descriptive scales. please make your judgements on the baSlS 0 mean to you. ent item to be judged and beneath h item on each scale in order. On each page you will find a differ it a set of scales. Please rate eac Here MICHIGAN STATE UNIVERSITY RESEARCH Michigan State University East Lansing, Michigan Consumer Shopping Habits in the Detroit Area Phllo, I'm ______ from Michigan State University Research. We are conducting a survey of shopping habits in your area. I would like you to help us by answering some questions. 1.1. Does any member of your immediate family work for a department store in the Detroit area? .1 [7 Yes .2. 17 No. .3 17 D/K, N/A If gg continue interview. If Yes discontinue inter— view. Explain the purpose of the study and thank the individual for his assistance. This is an in— complete interview. Interviewer should consult instructions for locating additional subjects. INSTRUCTIONS FOR ANSWERING THE SEMANTIC DIFFERENTIAL The purpose of this study is to measure the meanings of certain things to various people. You will be asked to judge these things against a series of descriptive scales. In giving your answers, Please make your judgements on the basis of what these things mean to you. On each page you will find a different item to be judged and beneath it a set of scales. Please rate each item on each scale in order. Here are some examples of how to use these scales: If the item to be judged was city and the scale was big — small it would look like this City :Small o Big: on w n u on so to n u n 0 Very Quite SlightlyzNeutral Slightly Quite Very If you think city is very small you would mark the scale as follows = 3 X =Snall Slightly: Quite : Very - a o o c 1 Very : Quite 2SlightlygNeutral o a o o o If you think city is §1ight1y big you would mark the scale as follows - X :Small In an Very Quite :Slightly‘Neutral 'Slightly' Quite Very If you think city is neutral, equally applicable to big or small, or that the scale big - small is not applicable to city, you would mark the scale as fOIIOWS: ° : :Sma11 - X . Slightly: Quite = Very 1 o o c e vs SlightlyiNeutral Biq: : : Very SEMANTIC DIFFERENTIAL INSTRUCTIONS IMPORTANT Please place your marks in the middle of the scale spaces, not on the boundaries. This Not this Never put more than one mark on a single scale. Be sure to mark every scale for every item — do not omit any. Please move directly from one scale to the next — do not check back and forth. Don't worry over the items. It is your first impression that is important. On the other hand, try not to be careless, we value your judgement. Please try to fill in the scales at a fairly high speed. ‘ 1. 1h .1. 1.. labon one-y my M- I nube' t—nt etote- 1n the Detrolt area. Please ell Ile- you lurk the following scales. l-l- iley shoe: 2 : : : , . -They the Invest styles: v : ‘ : - . - "_“."' m then other ery Quite Slightly‘leutrll -Sl1qhtly- Qulte : vary :the hart-1t store- l.2. Selle loo wellty: : 1 = g , , I-Idlndilh Very = out. =Slightly=ueutral :slxqmly= Quite = Very 1.3. The-x chi-Ind : ~ they sell 1- th;: v.ty : augg. - kind I like to 1.0. You can never {1 : : : z z : : ride variety of: Very : quite :Slightly.fleutral :slightly: quite : Very : It;ings to choou 1.5. A : : z : z z : place to get nee: Very : Quite :Slightlygfleutrel :slightly: Quite : Very 1deae ehouth furnishing- l.6. larely rune mt : : z : : : . ”f rchandiae: Very : Quite :slightlyzfleutral :Slightly: Quite : Very :w 2.1. Very strict in: = : : : : : e = = Quite =slightly=leutral =Sliqhtly: Ouite : Very 2.2. : = : = Quite Slightly; neutral “Slightly‘ Quite = Very 3.1. : : : z : you feel that you: Very : qu1te :slightly:neutral Slightly Quite Very “n. 3.2. Sales peopk : : : 1 m1: larch-ndile: Very : Quite :Slightly Neutral :Slightly. Quite: very wry 11 1.3. Sal I people : 2________t_______:________:_______J not interested 1n: Very : Quite :Slightly:fleutral :Slightly: Quite : Very being of oerv to 3.4. sales people are: : :____:_____:_—_:____:__. Very : Quite :Slightly:fleutral :Slightly: Quite : Very customer- ).5. Sales people Ink —-__:—:__:—_:—_:—_- feel uncoaidrt Very = Quite :Slightly:leutral :Slightly: Quite : Very able if you are just browsing 6.1. IV or : : :______:____:____ sale. the sale: very : Quito :slightly:Neutral :Slightly: Quite . Very nrchandi la eqular quality 4.1. Prices are 11k e):y : : :___=__=__—.=.___-P to be higher: Very : Quite :sliqhtly:leutral :Sliqhtly: Quite : Very here th othzr depu mnt roe (or the mh ndiae think of them newes st styles later the an othe:s department stor ell: highe quality zmerchandi he merchand lie they sell is not the kind I like to buy can al:ays find a H1 varietye of0 (things t hoos oor to get new idea s about horn furnishing 5 :place -Frequent 1y runs of merchandise Very liberal in - wcaeh 1ng cus— tme rs‘ checks -Very hard to get char arge accoun t Sales people make .you feel that you are impor -Sales people don't :know their mer- chandise very well ales people are zintereste d in being of serv ce to WC Sales people are :discourteou us to custm .Sales people make :you feel comfort— able if just browsing When they have a ale e. be ular quality rices are [likely -to be low here tha ein other department sto ores for t e same merchandise AAAAAZZ/ 122m m m m m7 may £22227 [ZZZZZ7 [22223 [222227 @2223 [ZZZZZ7 (ZZZZF A warm fr1end . ‘ store: very : Qulte The store is often: too crowded or: enjoyable shopping Oulte 5.3. The way the: departments in th e; store are arranged makes ‘1 e sy to (in things Very : Quite : 5.4. I an unlikely to: meet friends in: t are Very Quite endly to: - colored people: Y to see uppe pe 5.7. You are likely to: : see middle class: ple there 5.8. cu are unlikely: to see plower class: eople the ere Very Quite _ H w = r, In 1: < < (I H .< ny shoppingm :here You can bel ieve : their: advertisements S ... ,. «0 Very 7,1_ They are strict: on exchanges and: returns very Very Quite e a r I purchase them Hard for me to: ‘39t t re : 8. 2. Parking: faculties: are excellent Very lnconvenlent. : because 1! 1* not: near other 5‘ area n1 to shop at - ~ : : :sliqhtlytueucral slightly: quite = :A Aeold unfriendly :otore very :The storeie never li9ht1)’. Quite Very :too ore-dad for enjoyable shopping ~ - : : :fhe Hey Slightly:Ieutra1 :sliqhtly: Quite : Very : r nts 1n the store are rr es 1 hard to find th1nqs : :I a- likely to Quite : Very :meet friends in th : ::unfr1endly to .Sliqhtly: ~1Ieut.ral :Sliqhtly- Quite : Very :colored people : :You are likely Slightly: ~lleutral :Slightly- Ouite : Very :to see upper close le ther : : z are unlikely to Slightly: leutral :Sliqhtly; Quite : Very 5... middle cl 3 people t e .You are likely to Slightly_Neutral ,Slightly_ Quite , Very ,m loner cla :- people there — — :—:____.—‘_-1 can save Slightly: Neutral :Sliqhtly: Quite : Very z: shopping there : :You can't believe Slightly: Neutral :Sllghtly: Quite : Very :t eir advertisements ___... ___ -—;__.__. ey are liberal Slightly- -Neutral -Sllghtly: Quite ; Very Aon exchanq s and n lightly. Quite Very purchase them —:—:—:__: sy for lee to Slightlymeutral :Sliqhtly: Quite ' Very =:et the — _1 - : :Parking .Sliqhtly.Neutral .Sliqhtly. Quite Very :facllities are poor : :Convenlent .Sllqhtly: Neutral :Sllqhtly: Quite Very -beceu it I want to shop a' my am am am ”I he s n‘s: of w stereo in the Detroit area. Please think or th- all when you I I———-—-:~———' —‘ ” “mm '2:- V-rv . out. -su¢xuyxs—tul sugar out. :the nmer styles hart-It stores later than other department stores :Sel ls high qual 1ty fisherchandi l.1. sells it- guilty: misc sex v." mite ‘slithly: leutral :sliqhtly3 Quite l.) The —r 1ss::___.____ __ __ .—-.__ ______ :The merchandise "‘7 'f“ h "m Very = min Slightly-Neutral-anr~t1y= Quite 2 Very :they sell is not kind I like to My the kind I like [0 1.6. You cs- never hnd. . : . . ou can always s wide variety of: Very = quite :Slightl Neutral :sliqhtly: Quite : Very :find a H1 1de things to chooee variety of things from to Choose {r 1.5. A good: : : : A poor place to get now: Very ; mite :sliqhtly: Ieutral :Slightly, Quite : Very :place to get new ideas shout ho- ideas about home furnishings furnishings 1.6. Rarely runs out: . . . :Prequently run of merchandise: Very : mite Very :out of Iner rchandlse 1.l. Very etrict '1.- : 1 :Very liberal in c hing one Very = mite Very cashing cus— t checks 1 : :____: Verya hard to get a dug- setoant Very = mite 3 Very ii 1’99 ICCO 0“!“ Ll __ ___ ____ _____ ____,____ e e m“. quite : Very :you feel that you are important 3.2. salespeople knot: : 1 sales people don't their urchandiu- Very Quite .Slightly: neutral lightly: Quite : Very :knov than mu- vor 1 ll chandise Very well 3.3. . Sales peopl soles : : = _.__=_ __ not interested in: Very : ouite .SlightlyJOeull’ai lightly: quite : very :interested9 in being bi t to custa-rs customers Sale: people are SJ. Sales people are: v z : : : : courteous to: wry .- Quite :sliqhtly:fleutrsl slightly: Quite Very COUI::0 u- t0 cust_rs custome 1.5. Isles people n‘s: - Sales people make was {—1 mint—z Very : Quite :islightly neutral .sliqhtly quite Very :you feel cry-(ort— file it you as able 1! you are just ing just brow-ing 0.1. they hues s: - : : : -. when they have a sale. the sale: Very mite :sughtlyzseutral lightly: Very ale. the u all merchandise is be— .reh—liee is or law re gul or quality regular polity 6.1. Prices are likely: x : z 2 : :pnc" ":r likely to be higher: very : quite ightlleeutrsl :Sllghtly: Quits : Very :to be 101! here tb- i- here than2 1 Newest-Ir statutorthe C.- rchandlse £222? 5.2. 5.4. 5.6. 5.8. 1 Harm fr1endly: $10!. 2 Quite : : : : I =Slightlyzleutral =Slightly: (31th x Vsry tk cold uh i-Oly IIIIIII ore to! Very The store is often: : 2:17. Itorois saver m loaded for- Very : Quite :Slightly_ neutral :slifiltly: Quite : very :tooc enjoyable shopping sajoyible “sizing Th e Hay the e: : : - : z I :‘I'h- the m departments in he: very : Quits :slightly:hutral :slightly: (rite Very : nts 1n the are are arranged store are xr makes it easy to do it hora to find things find. things i an unlikely to. -1 an likely to m meet friends in: Very Quite :Slightly: leutral :Slightly: mite very :Ieet friends in the store the store nendly to.____ __ __ .___ —.—— ____.::____: Unfriendly to m colored people: Very Quite : Very :colored people Y u re unlikely. : : :You are likely m to see upper class: very Quite ' : wry :to see upper class people there people there You are likely to : : :You are unlikely to m see middle class. Very Quite ; Very :see middle class pea 9 there people teh set are unlikely: : : on are likely to my to see lower class: Very Very .see lower class eople there people there 1 can't save: : - :I can save m any time by: Very quite -Slightly: neutral :Slightly: Quite : Very :ti-e by shopping there shopping there You can believe. : : :You can 't believe m 2 Very Quite :Slightly-: Neutral :slightly: Quite Very their advertisements advertisennts They are stric ___:_____: :____x_______:____ They: are Iliberll m on exchange: and: Very : Quits :Slightlyzueutral :Slightly: quite Very :on exchange sang returns re etu ay ———""—————f———.————’ m satisfied with Very Quite :Slightly: Neutral :Slightly: Quite . ‘Iery :latisfied with e goods In the goods 1 buy there at er I there after I purchase them purchase them ard for Is to . z __ :Basy fore as to m get the Quito -Slightly eutrsl .Slightly:ou1te -tqet Parking : Parking m7 facilities: Very : quite .Slightl eutral :Sllghtly- mite : Vsry :(scilitlel are excellent are poor In convenien : xConvenient m because it is not: Very Quite -slightly: Neutral :slightly- Quite Very :beclu it is near other stores “In sahop It ‘ 3. We]. Depart-ent Stores ~ Mala has a nulber of stores in the Detroit area. Please think or them all when you mark the following males. 1.1. . . . . . , , th They Shw"‘,¥'——'——"-—.‘=\=Th Sho 5,1, A warm fnendly. ; : ‘ :A cold unfriendly m e newest stylem Very : ou1te =sliqht1y:Neutra1 :sughtiy: Quite : yet :the newest st-qes .—-_ wh— —1—-1—‘h_tT T ——_'3t°re sooner than other y lat th ) Store Very . Quite 511g t1y-Neutra 5 1g y ou e ery . department stores er an Other department stores 5 2. The tstore 15 often: : : : : :The storeia never ZZZZZZZ7 w d f0 or: Very : Quite :Slightly; Neutral :Slightly; Quite ; Very -too crowded for 1-2- $2115 10" qua . - : : Sells high equality m . enjoygble shopping enjoyable shopping {chand:::~ Ver Quite -Sliqhtl. Neutral islightl Ouite f Verv =merchandis Y Y Y . . m 5.3. The way the:__ ____ ~:—:—: _—:___: The way the m 1-3- The “‘"Chandlse .— —_.=_s=‘_‘ =—_=%‘The merchafi‘dlse departments in he- Very Quite .511ght1y.Neutra1 :Slightly- Quite : Very -departments in the thevs sell is the: Very Quite ~s1ight1y=ueutra1 :slight1y= Oulte = Very :t‘xey ell A: not 5m .3 are arranged store are arranged kind I like to buy t e kind 1 like makes 1t eas to make 25 it har g to buy things find thi 1.4. You can never find: : :You can always m 5 4 I am unlikely to~ _ . y ‘ ' ‘ -I am likely to m a “’de ”new ° ' 'Shghtly' ”“1” ‘ V9” find a “de meet friends in: Very : Quite :Slightly:Neutral :Slightly: Quite : Very :meet friends in th1ngs to choose varIEtY 05 things the store the store fr to choose from 5.5. Pr endly to; : : :Unfriendly to m 1.5. Agood. : r A poor [Z/ZZZZ/ h_———*——* —— .— —.—— — *‘ —— .4 colored eo 1e: Ver Quite .Sli htl Neutral .Sli htl ~ Quite : Ver :colored 0 1e place to geth new: . . lightly: ouite : Very place to g t new P p y g y g Y y Pe P ideas about ideas about home 5““‘5hin‘l‘ ““15““‘35 5 e. 011 are unlikely-_ : : : :You are like m to See upper class: Very : Quite ; Slightly, Quite : Very :to see upper 1class 1.6. Rarely runs out: 1 : : : : .Frequently run m 9901”” the” peeple n‘ of merchandise: very : Quite :Sliqhtly:Neutral :sliqhtly: Quite : Very _ ut of me rchandlse 5.7. You are likely to: : : : ~ : : :oY u are unlikely to m 2_1. Very sag“ in. . , z . . ery liberal m —— see middle Class: Very : Quite :slthtlymeutral :Sllqhtlv: Quite : very -see middle class cashin cus—: very : Quite =Sliqht1y=Neutral =Slightly= nu1te Very lcashin ens- People Lhere peop1e there tomers‘ checks t mers' checks 5.8. You are unlikely: : z : :You are likely to m 2.2. very easy to get: :___:__: —:___: ‘z‘:Very hard to get m to see lowe er Class: Very ; Quite :Sliqhtly: Neutral Slightly: Quite : Very :see lower class a charge account: Very : quite =Slightly= Neutral =Slightly= Quite = Very =a Charqe account ple there people there 3.1. Salespeople make: : : : : : 2 a s people make AZZZZZU you feel that you: very : Quite :slightly:Neutral =slight1y: Quite Very :you feel that yOu 5 9 I can't sav . . . can save my im a ' ' H—'_‘~— *‘h' _ .— an “import“ a P I an me by: Very : Quite :sliqhtly: Neutral :Sliqhtly: Quite very :t me ’ slxzpplnq there shopping there 3.2. Sales p:opl e know. 1— ~ ~_ : ales people don t [$227 the eir rchandise; Very .Slightly.Neutral .s1ight1y. Quite : Very :know their er— 5.1. You cm believe: _ . . You can. t believe W W we” c handise very we“ their: Very : Quite . .sliqht1y~ Quite : Very :rheir advertisements advertisements 3.3. Sales people are: : : H—z— : :Sales people are «m not interested in: Very : Quite :Slightly: Neutral :Slightly: ouite : Very :interested in being 7 1 They are strict- - - : : : They are liberal [22227 be Of service f sex—Vice to on exchanges and: Very : Quite :Slightly: Neutral :Sliqhtly. Quite ; Very :on exchanges and cu: stome custcmers ret Ins returns 3.4. Sales people are: : —=.__7__=1._. _~S le ople are m 7.2. I am a] ways- . . . 1 1 am never m courteous to: Very Quite .s1ight1y.Neutra1 .s1ight1y: Quite = very rgiscourlzous to sat1sf1ed with- Very r Quite :Slightly: Neutral :Slightly: ouite ;_ Very :satlsfled with we” the goods 1 buy the goods 1 buy there after 1 there after ‘ [Ch 3.5. Sales peopl make —_-W:____Z—:—__:‘:_“:Sales people make m PUIChase them purchase em you :eel uncomfo very : Quite :SllghtlytNeutral :Slightly: Quite : Very :you feel Comfort- if you at e aist ginylou are 8.1. Hard for me to- _ = z : Easy for me to 1222227 . 5t browsing 1 . 517K] get ther re : Very Quite :slightly: Neutral :slightly. Quite : Very rqet there 4.1. when they have a: :When they have a LZZZZZZ7 , . h_—,e— E 8.2. Parklnq- - 2 : :Parki ng [222m sale, the 9312- Very Quite iightly. very :sale. LkA? sale :‘VE' — W——-——h—— , merchandi is merchandise 15 be aref::c:::::: Very . Quite :sliqhtly: Neutral .slightly. Quite . Very .f::11it:es regular :ualizy 1uw reguiar quality 4.2. Prices are 1‘1 -. . : _ =—1‘=m Prices are r11ke1 , 8.3. Incon'lenlent- : : 2 ‘ : :Convenient 122227 —— —~._ —— . . Y LZZZZZZ7 — —- — — _ — E to ehiqher- Very : Quite :slightlyzNeutral =5119htly= Quite = Very . o be 10 _ because it 15 no Very Quite :slightly:Neutral :s1ight1y. Quite Very :because it is herb re tha here thane in near other stores near other stores other departmen: other department 1 want to shop at 1 want to shop at stores for the stores for “1 Blue merchandise same merchandise ‘ Thc'ldcar Department Store If were thinking utoro does not have to exi 1-1- The ey Show. the newest styles: sooner than other department stores Very Sells low qua al1 chandise. - The merchand 1 se - they sell 1s 1 kind I 11kg to buy You can never find: a wide var1ety of: thlnqs to choose from Very A good: place to get new: ideas fabout home ishlnqs Rarely runs out of merchandlse Very Very str1ct 1 rushing cus tomers' checks Very a charge account: Very Sale: people mak you feel that yo . are unimportant very Sales people kno the eir merchandise: well Very sales people are: not interested in: ing of ser rv1cv to rustomers Very 3.4. Sales copeople ar eons t atelier. 3.5. Sale es people make: vou feel uncmfort abl if you are Jus sl brow-ing 4.1. when theyh have seal 33 Ill-r rach ndile is reqular quality 4.2. Prices are like to b e hiqher-# 1n here than other department stores for the same merchandise Very Very Very Very of the 'ideal' department sto st; it Simply is Quite Quite Oulte Quite Quite Quite Unite Quite Quite Quite Quite Quite Quite Oulle Qulte how would what you imagine would =sl1qhtly=Neutral .11qh11y: llghll; _: : cold unfrierr‘Y de rtmen Very Quite lightlygueutral :Slightly: Qu1te : Very :departme nt 5.2. This department 1 ~ ; -T This departmen 1 always too crowded: Very quite lightly. Neutral :Slightlv: Quite : '.'ery :never too owdeu for enjoyable for enjoyable opping shopp 5.5. Friendly to: - : :Unfriendly 1c colored people: Very : Quite :S‘l whtl liqhtl Quite : very colored people- =___Vi__—: You are un n11 I 3 :. .Slithly. Quite Very xsee upper “.295 eople the 5.6. You are likely to: see upper class: Very people there Neutral Quite : Slight! u 7. You are unlikely: : - -You are likely to to see middle class: Very : Quite Slightly: Neutral :51. qhtl Quite : Very :see middler clan- people there people the ‘ .H You are likely: : :You are unlikely to see 10.! uer Class: Very : Quite :sl1qhtly: Neutral Slight] Quite Very :to see lower clue- ma an there gole her. 549, 1 can save __ by s .oppinq there: Very :1 an't save tune :hv shopping there . Quite :EingymeutraI :51iofiTy:_ou1re :‘7 av—T ewfiau/ .ZZZZ’ I 9. More lbn'e clothing Depart-ht (nu. rnIpo-ndente only) ‘ . lo. Mem'e lath-re Depart—It (Kale relmnte ally) 'leue ml: the (allowing Icelee .- you feel t becrihe Indem' len' 0th eer mm aura“ em: 1 2 hey I l c1 ing nepert—u. Pleeee k the (allowing Icelee II you 20.1 they heartb- Ion' I re pep-r - - I'll. 1°" «Inky: a x x x x x 13.11. high «.111, m 1,1_ , men. '“nlgn quality Meet very x mite nllightlyxuutnl xsughtly: Quite 1 Very x-rchendile 1cm Very 2 cult. :suqhtly lean-1 xsllqhtly: who r Vary m1 1.3. n. A “ 1 . g x : 1 mm lurchumlIe m 1.3. rch-mun '1‘" "c ‘5'! eel ie the: Vety ! mite Islightlyxleutrel lsllqhtlyx Quite I very xthey Iell iI not they :11 1It1he :t u 1 u not “ll 1 10!- to h! the kind 1 11k. kind 1 11):- 0 buy a" “"4 1 “h to buy to buy I _ _ quu can Il any 14. You e never: 1 x x x z 1 You can I1 we m =“M I '14. variety the I ue- mietyx Very : mite ISlightlyzleutrel :sughtly: Quite 1 Very :tinde wide v:riety of thingI to choeee 0* Bil-n9- t m o! thinge to choo- trc- "°‘ Ira 1.5. luely=_—=__X__ _=_ __ —: Frequent lv Ll. may: I 1 : 1 = x :Proqutntlv m n," am wry 1 mite .Sliqhtly.lzut 119mm. . Very =mm run at: Very . Ouite :Sliqhtlyxleutre . Quite : Very :mn- Wt of _xcmu. of Ierchlndiee o! of nercMndiIe . 3. 2. kSale: people don' t-____ __: :s-leI people know 3.1. Selee people Cu: 1 x : 215.13. people now am new their eerchAn-: Quite . .sliqhtly :their heron-Mite k heir urchen-z Very : Glit- 1Slight1yzleutrel xSlighlly quite Very their merchnndi-e ‘31.. v.” “u very diee very hell verywe wel 3 5 c1.x‘.-n. 3.5. Clerk. lake you: : --____ _- _.__. ___=C1"x' “I" M ' ' —— —— — __- __ ==__'— Clerkt Bake e,you 22222227 f 1 g t m , v = t 11 m1 neutral -s11 htl . Quite : Ver :feel unco-Ioruble (eel cuzfiteblex Very 1 Quite :sliqhtly: Neutral .slightly: mite : very :fe e1 uncomfortable .; yzrf: ;u . ery Gui . 9 Y 9 V V I V0“ " n 11m uju “I: if you c qut um, bro-vein :xr ing brouei 4.1. W 4.1. when they have 1: : — __ =___=_="h'" ““1! “'V‘ ' vex-y : Quite :Slightly:leutra1 .Slightly. n “1 , the lelex Very = Quit- .sughtlymeutnl :s11qm1y: ouite = Very :--1- th- "10 ‘ .b._ merchandise iI of nrchendi- ie be- Owlu quality 1w “9111:: quality "9““ ““1“" 1°" I ll: “.1“, LI. Prime Ire likely: - : - ~ ' _, —..‘— ~-.—__ __= :Prxcel are likely 4 2 p 1“ . - :vri ee are likel Dd 11.3122" :2- Very x Quite :Slightlymeutr-l :Slightly. Quite : Very .to be lower at m ' ’::°;e";:qhu°_{.wim. mim.mi cm:- ; Very .to i. 1a...y the“? HudIon'I than in udlon'I then in udIon'I th in ItoreI for the other department at er deper nt other depart-Int d1“ ltoreI for he I oreI for the to I -' one merchendiu um. nerchandi-e II— urchendiee 5.1. A were - ~ __ ___ _— friglyx Very :W;m:m;W=:—mr ery 1A cold unfriendly m 5,1, A .,“-m friendly: 2 : cold unfriendly “P“ 1 x U deputme de p-artnent ley z 0111:. slightly: neutrIl -Sliqhtly. Quite : Very -deput Ien 5. 2. Thie “pert-nut i '1‘“?! .! - - . . “m "M, v", : quite :B—liqhtl —_uuu «W- ‘ . x dtpextnent 1. m 5.2. ‘l'hiI depart-en: 1 = . 2 rh-1- deparE-nt 1- «enjoyed. ’ ‘1 Y= 9“ a = V")! :never too crowded alwayl too c owded Very : quite -sught1y:Ieutn ' : Quite 1 Very :never too round . ing for enjoy-ble for enjoynhle for enjoyeble M” Ih°1=pin9 unoppmg Ihwpinq 5.5. Mind 1 x x 1 1 __ 1 mntri 61 1 colored “1:. Very l mu. [slightlyxleutrel xSlightly, 0“" if. I‘Very‘ mole W: Y 0 m 5.5. endly to: .Untriendly to Pr colored peoplex Very x mite xsliqhtl .neutrel -Sliqhtly. Very . olored people 5.6. You Ire likely to: a l x t x _— — — __ 1 You It. \mllk:ly to m 5,5_ ., lik 1 t . . 1111 1 t women-r1. cit-:1 Very : out. asliqhtly zleutrel muqhuyx aux:- . Very :Iee upper °“ '" 'c"_ ° ——v"y :—-——°““. 1bliqhtly:leutrel :Sliqhtly- mn- ' I”: 3;;“nch ' ’ ° people there- ople there people there 6.1. You ere ulfelyx 1 x x 1 1 _— .— _— _' eeeliule 1m: v.17 . ouite :llightlynleutrel xsliqhtlyx no“ '" uh” to m 6‘7 w“ '" “n“ V‘ “1.3.1. mite 1 very :Iee Iiddle cleee people the —- _— -___ ____x __ 2 11"» are likely to to Iee Iiddle cl:II- Very 1 Quite usllqhtly: Ieutrel :slightly; Quite 1 Very uee Iiddle clIII ere people there 5... You ere likely; I x x x x m u... lot-u 131..., M .mite alightlytmtzel xSlightlyx quite : very xYou u. “flux.” 5'8 0 I" I'm-1y: e-le then to Iee 1m Mt cl I mmle there: :J. l m ..‘ tile‘ : - 1 = 1 1 c“. ‘ s 9 1 cu: Ieve ti-X ' - v _T__.____._— ..n ti- - - __."__—_~___.~_-_ _. X1 cen't Ieve ti- by 1110.191“ on“: Very : out. .Sliqhtlylleutral :Sliqhtly: Quite : Very my flapping a.“ @2227 by “099139 there: "I! ! 0°"- miahtlw'wtnl =SU9M1V= oult- = VH1! :11» nhmmq thor- -‘_‘___ _._. _._.____.I _ I____I IYou Ire “11111.1, xto IeI lower cleII very I quite galightlyxleutrel xslightlyx Quite 1 Very 1ta Iee lover ole-I 1. there 'fleoole there - ‘ ‘1? I _ 11. Shopping Please think of the activity of shopping -- from the time you leave until you return. Mark the scales to describe how you feel about shopping. Please disregard grocery shopping in marking these scales. For me For me 5.1. shopping: : : : : 2 ; :shopping 12222227 is fun: Very : Quite :SlightlyzNeutral :Slightly: Quite . Very :18 work 2 :For me, {222{{{/ 5.2. For me,: : : : . . shOpping: Very ; Quite :Slightly;Neutral :Slightly: Quite : Very :Shopping is dull is exciting 5.3. I like to: : : : z : _ : :1 like to [/22222/ shop alone; Very ; Quite :Slightly:Neutral :Slightly: Quite : Very :have company when I shop 5.4. Shopping: : : ' : : :Shopping 2222222/ Very :gives me no satisfaction gives me: Very : Quite :Slightly; Neutral :Slightly: Quite satisfaction :I feel 2222222/ Very :creative when I go shOpping 5-5. I don't feel: : : : . :Slightly:Neutra1 :Slightly: Quite creative when: Very Quite I go shOpping 12.1 What is yOur favorite department store? 222222/ 1. 27 2. 27 3. 27 . 27 . 27 . 27 2. If you are not able to find what you want at your favorite department store, which store would you try next? ZZZZZZZ7 1.52.53.27 .27 .5 .27 3. Which department store would you try after that? 22222227 1.27 2-27 3-27 27 27 27..- 4. If Hudson's is not mentioned above — Have you ever shopped at Hudsons? [2227 1. Yes 27 2. N027 3. D/K, N/A, etc., 27 (If no, omit questions 13, 14 and 15) 13.1 Which Hudson's store have you shopped at most often?_______ 222222227 l.DT27 2.115 3.E27 4.w275.D27 6. LP 27 7. p 27 8. M 27 9. OM 27 2. When did you last shop at a J.L. Hudson store? 2227 1. less than a week ago 27' 2. more than a week but less than a month ago{27 3. more than a month ago 27 3- Could you estimate how many times a month, on the average, during this past year, yOu shopped at Hudsons? El 1. once a month 27 2. two or three times a month 27 3. more than three times a month 27 4- Could you estimate how much you spent at Hudsons last year? [22227 1. less than $100 27 2. $100—$249 27 3. $250-$499 5 4. $500—$1,ooo 27 5. more than $1,000 27 14-1. How long have you been a Hudson's customer? 22227 2. more than 1 year and less 1. less than one year 27 than 5 years 27 3. more than 5 years and less than 10 years 27 4. more than 10 years 27 2. From what source did you first hear about Hudsons? 222227 1. from some other person 27 2. from an advertisement 27 discovered Hudsons myself while driving, or walking around town 27 other sources—please specify 27 don't recall 27 DJ Ulb 16.1. 17.1. When you buy something at Hudsons, do you 1. pay by cash or check? .1 always 27' .2 sometimes 27 .3 never 27 2. use your charge account? .l always 27 .2 sgmetimes 27 .3 never 2/ When you buy something from Hudsons do you purchase 1. in person? .1 always 27 2. sometimes 27 3. never 27 2. by mail? .1 always 27 2. sometimes 27 3. never 27 3. by telephone? .1 always 27 2. sometimes 27 3. never 27 When you are shopping at Hudsons, do you 1. shop upstairs? °l always 27 .2 sometimes 27 .3 never 27 2. shop downstairs? (budget store) .1 always 27 .2 sometimes 27 .3 never 27 When you shop at Hudsons, do you travel by 1. car? 01 always 27‘ .2 sometimes 27' .3 never 27 2. walk? .1 always 27 .2 sometimes 27 .3 never 27 3. public transportation? .1 always 27 .2 sometimes 27‘ .3 never 27 The J.L. Hudson Company advertises in the daily and Sunday newspapers. Do you read these advertisements? 1. always 27 2. sometimes 2/ 3. never 2/ How do your friends feel about Hudsons? l. the same as you do 27 2. differently than you do 27 3. D/K, N/A. etc. 27 Demographic Data 18.1. 19.1. 20.1. 21.1. Respondent‘s marital status 1. single 27 2. married 27’ 3. widowed 27 4. divorced/separated 27 4. other 27 Do you have any children living at home (include any children supported by parents who are away at school)? 10 yes 27 how many? 2. no 27 __’—- 1111/ A? Please indicate which letter corresponds to your age categOrY° A.27 13.27 0.5 13.27 E.,27 13.27 'Hand respondent card 20.1. What job does the male head of the family do? (If there is no male head, please indicate the job of the female head of the family.) ZZZZF 22.1. 23.1. 24.1. Thank the information you have given Please indicate which letter corresponds to the job of the male head of this household. Hand respondent card 21.2. ZZZéLLQQQ 1.17 2.27 3.27 4.27 5.27 6.27 7.27 8.27 9.27 If head of household is female, which letter corresponds to the female head of this household— hand respondent card 21.2. [222222227 1-27 2.27 3.27 4.27 5.27 6.27 7.27 8.27 9.27 (note: mark only 21.2 or 21.3, not both) Would you please describe more fully the job indicated above in 21.2 or 21.3. 1. Is university training required for this job? 1. yes 27 2. no 27 2. Is an apprenticeship required for this job? 1. yes 27 2. no 27 3. Does this job include the supervision of other peOple? 1. yes 27 2. no 27 BEI E1 Please indicate which letter corresponds to your total family income before taxes last year. Hand respondent card 22.1. A-27B-27C-JD-A7E-27F-[7G-z7 ZZZZZZE Please indicate which letter corresponds to the source of the greatest part of your family income. Hand respondent card 23.1. A.27B.27C.27D.27E.27 AZZZZF Please indicate which letter corresponds to the last year of school completed by male head of family. Hand respondent card 24.1. A.27B.27C.27D.27E-27F-27G-Z7 W Do yOu own or rent this dwelling? l. own 27 2. rent 27 [27 .1 If owned, ask: How much do you think you could get for this house if it were put up for sale today? Hand respondent card 25.1.1. . A.27B.27C.27D.27E.27F.27 ZZZZZZZZZZ7{ G.27H.271.27J.27 .2 If rented ask: What is the monthly rent? ‘ Hand respondent card 25.1.2. A. 27 B. 27 C. 27 D. 27 E. 27 F- [7 [222222227 g. 27 H. 27 I. 27 you very much for your COOperation. I can assure you that all to us will remain confidential. We aPpreCiate your assistance in this research. g The interviewer will complete the following by observation after the interview has been completed. Respondent's address Respondent's code number Respondent's dwelling type. .1 single 27 .2 attached one side 27 .3 attached two sides 27 .4 two family dwelling 27 .5 du-plex 27 .6 four—plex 27 .7 apartment 27 .8 other 27 specify Respondent's sex 1. male 27 2. female 27 Respondent‘s race 1. white 27 2. negro 27 3. other 27 specify 122222227 W W fl 2227 s‘ i 20.1 21.2 and 21.3 22.1 23.1 RESPONDENT CATEGORY CARDS Please indicate which letter corresponds to your age category. Less than 20 years More than 20 but less than 30 years More than 30 but less than 40 years More than 40 but less than 50 years More than 50 but less than 65 years 65 years and older. 'idL'IlUOUJlD Please indicate which letter corresponds to the job of the male (female) head of this household. A Professionals and proprietors of large businesses B Semi—professionals and smaller officials of large businesses Clerks and kindred workers Skilled workers Proprietors of small businesses Semi—skilled workers Unskilled workers QWWUO Please indicate which letter corresponds to your total family income before taxes last year. Less than $4,000 $9,000 - $7,999 $8,000 — $9,999 $10,000 — $12,499 $12,500 — $14,999 $15,000 — $2u,999 $25,000 and more Q'filthlUOKIlil> Please indicate which letter corresponds to the source of the greatest part of your family income. Inherited wealth Profits and dividends Earned salary Earned hourly wage Other — please specify [TJUOUSHD 24.1 25.1 25.2 238 Please indicate which letter corresponds to the last year of school completed by the male (female) head of this household. WINDOW?> Grade school or less Some high school Graduated from high school Some college Graduated from college Graduate or advanced degree How much do you think you could get for this house if it were put up for sale today? (TJ'IILTJUOUJS> More than $50,000 $35,000 — $99,999 $20,000 — $34,999 $15,000 - $19,999 $12,500 — $14,999 $10,000 — $12,499 Less than $10,000 What is the monthly rent? QWF‘JUOWID More than $500 $250 - $499 $150 — $249 $100 _ $149 $75 — $99 $50 — $74 Less than $50 APPENDIX F INSTRUCTIONS FOR INTERVIEWERS 240 Instructions for Interviewers The following instructions are a guide to the administration of your interviews. In some cases, such as dress and time, they will be of a general nature. In other cases such as the termination of interviews and the selection of additional subjects they will be specific. General Instructions gees Interviewers should wear a jacket and tie, or Similar appropriate female attire when interviewing. In addition, the identification card should be worn in a conspicuous position. Both dress and identification will increase the interviewer‘s effectiveness in gaining subject cooperation. iiine All interviews should take place before 9:00 p.m. Car should be taken to avoid meal hours. No interviewing may be done on Sundays. To make the best use of your time, you should arrange to interview when a number of interviews can be done at one time. To facilitate this, your subjects are located close together. 241 Interviewer Kit Each interviewer will be provided with a kit containing the following items: 1. Street map of Detroit; 2. List of addresses of subjects; 3. Identification card; 4. Instructions; 5. Set of respondent category cards. Specific Instructions When you go into the field to interview take your interviewer kit, a pen and pencil, and sufficient questionnaires. Male—Female Quota Out of each four households attempt to obtain an interview with one male head of the household. As this may not always be possible, please try to obtain interviews from twelve female and four male subjects out of your total of sixteen subjects. Introduction To Subject In your introduction to each subject begin with the first paragraph of the questionnaire. Offer as little in addition as possible. If you are asked how long the questionnaire will take, emphasize that it Will only take a short time. If you are pressed, tell the person that it will take about 25 - 30 minutes. 242 Question 1.1 The first question is set up to eliminate persons who are closely related to people employed by Detroit area department stores. For your purposes the Detroit area may be thought of as metropolitan Detroit. The following is a list of department stores: Arlan's, Atlantic Mills Thrift Centers, Crowley—Milner, Demery's, E. J. Korvette, Federal, Gem International, J. C. Penny, J. L. Hudson, Jacobsons, K-Mart, Miracle Mart, Montgomery Ward, Peoples Outfitting, Sams, Sears—Roebuck, Shopper's Fair, Spartan, Topps and W. T. Grant. To the people who answer Yes to Question 1.1., explain that you are investigating the way people feel about department stores and that it would be unfair to ask them about these stores because of their close relationship to one of these stores. Thank them for their assistance. To people who answer No to Question 1.1., explain that it would be easier to complete the questionnaire if you could sit down with respondent. Try to gain admittance to the house and sit in a place where you can assist the subject. Semantic Differential Go over the instructions for answering the semantic differential with the person until you are sure he or she understands. Then, go on to the actual T’f semantic differential questions having the person fill 243 the scales in on his or her own. Do not offer any additional information; simply explain that it is their impressions you want. Have them go as quickly as possible from one scale to the next and one concept to the next. General Shopping Questions For the general shopping questions (12.1 to 17.1) the interviewer is to read the questions to the subject and record the answers in the appropriate places. In question 12.4 (and in four others) D/K represents "don't know,” N/A represents "no answer." In question 13.1, 1. DT represents Hudsons Downtown store; 2. N represents Hudsons Northland; 3. E represents Hudsons Eastland; U. W represents Hudsons Westland; 5. D represents Hudsons Dearborn; 6. LP represents Hudsons Lincoln Park; 7. P represents Hudsons Pontiac; 8. M represents Hudsons Madison, 9. OM represents Hudsons Oakland Mall. In questions 15.1 to 15.”, for each section of the question, mark only one of: always, sometimes or never. e.g., 15.1.1 pay by cash or check: 2AM 1. always 2. sometimes 3. never Demographic Data The interviewer should ask the subject the questions in the demographic section and fill in or mark the appropriate place. Questions 20.1 to 25.1.2 are to be answered with the use of the appropriate respondent card. The subject should give only the category letter not the actual information. Questions by Observation Questions 26.1 to 27.2 are to be filled in by the interviewer after he leaves the subject's home and before he begins his next interview. Scoring and Turning in Questionnaire On the far right edge of each sheet of the questionnaire you will see numbered boxes. These are for card punching and tabulation purposes. Before turning in the completed questionnaires please mark these boxes to correspond with the answers given by the subjects. Completed questionnaires should be turned in the following day to Mr. Peter Arkison. Spot Checks, As is normal procedure, spot telephone checks will be made of completed interviews to ensure authenticity 245 Obtaining Additional Subjects You may lose a subject for any one of a large number of reasons: 1. household address no longer exists; 2. subject refuses to cooperate; 3. after three call-backs you are not able to contact the subject; a. only wrong sex is available and you need to balance your quota of subjects (u male, 12 female). If for any of these reasons, or any other, you must add a subject, simply count two addresses along the same side of the street in the direction your subject address list goes. If you have reached a corner in your area (as shown by sketch) turn left and count two addresses. If the address picked for an additional subject is already on your subject list, count two more addresses until you find a new subject not on your list. In counting addresses to obtain additional sub— jects in multiple-dwelling buildings: 1. Count from the lowest to the highest floor (the lowest letter or number in a multiple dwelling is the lowest address for counting purposes). 2H6 2. In a house with a lower, an upper and rear residence the lower residence is the lowest number, the rear the highest. In order to facilitate the adding of additional subjects follow the list of subject addresses as it is given. Location of Subjects You may find that because of peculiarities of some neighborhoods the addresses in your list do not conform to the accompanying sketch. In such cases disregard the sketch and follow the given addresses. You may find in some cases that the address you have indicates the lower, upper or rear residence at a particular street number, but there does not appear to be more than one entrance. In such cases simply ask at the door for the resident who lives upstairs or at the rear depending on the circumstances. APPENDIX G IMAGE DIMENSIONS 2H7 ,' l. 248 This appendix presents the division of the thirty bi-polar semantic differential scales into four image di— mensions. Merchandise Suitability 1. Fashion innovativeness , 2- Quality of merchandise 3. Appeal of merchandise Variety of merchandise 5. Home furnishing ideas 6. Merchandise availability 7. Prices 8. Sale merchandise quality 9‘ Post transaction satis- faction Sales Personnel Mm Attitude of salesclerks They show the newest styles sooner (later) than other department stores. Sells high (low) quality merchandise. The merchandise they sell is (is not) kind I like to buy. You can always (never) find a wide variety of things to choose from. A good (poor) place to get new ideas about home furnishings. Rarely (frequently) runs out of merchandise. Prices are likely to be lower (higher) here than in other department stores for the same merchandise. When they have a sale, the sale merchandise is of (below) regular quality. I am always (never) satis— fied with the goods I buy there after I purchase them. Sales people make you feel that you are important (unimportant). 249 Knowledgability'of salesclerks Interest shown by sales— clerks Courtesy of salesclerks Salesclerk reaction to browsing Check cashing practice Return goods practice Store Congeniality /' l . Friendly atmosphere Congestion Likelihood of meeting friends Attitude toward Negroes Upper class store Middle class store Lower class store Trust in advertisements Charge account pOliCy Sales people know (don't know) their merchandise very well. Sales people are (are not) interested in being of ser— vice to customers. Sales people are courteous (discourteous) to customers. Sales people make you feel comfortable (uncomfortable) if you are just browsing. Very liberal (strict) in cashing customers' checks. They are liberal (strict) on exchanges and returns. A warm friendly (cold un— friendly)store. The store is never (often) too crowded for enjoyable shopping. I am likely (unlikely) to meet friends in the store. Friendly (unfriendly) to colored people. You see are likely (unlikely) to upper class people there. You see are likely (unlikely) to middle class people there. You are likely (unlikely) to see lower class people there. You can (can't) believe their advertisements. Very easy (hard) to get a charge account. 250 Locational Convenience 1. Ease of access 2. Parking , 3. Shopping efficiency , 4. Store layout 5. Convenience to other shops Easy (hard) for me to get there. Parking facilities are excellent (poor). I can (can't) save time by shopping there. The way the departments in the store are arranged makes it easy (hard) to find things. Convenient (inconvenient) because it is near other stores I want to shop at. APPENDIX H STATISTICAL PROCEDURES 251 252 Appendix H describes the statistical procedures, formulae and identities used in the application of the Wilcoxon Matched-Pairs Signed—Ranks Test and the Mann— Whitney U Test to the data collected for the research. Wilcoxon Matched—Pairs Signed-Ranks Test1 The Wilcoxon T is a nonparametric test to determine the probability that matched pairs of variables come from the same population. Procedure 1. For each matched pair (d , d ), determine the signed difference (di) between the two scores. 2. Rank the di's without respect to sign. With tied di's, assign the average of the tied ranks. 3. Affix to each rank the sign (+ or —) of the d which it represents. 4. Determine T = the smaller of the sums of like— signed ranks. 5. By counting, determine N = the total number of d's having a sign (pairs with a d = O are dropped from the analysis). 6. The procedure for determining the significance of the observed value of T depends on the size of N: a. If N is 25 or less, . . . (The Table of Critical Values of T in the Wilcoxon Matched- Pairs Signed—Ranks Test2) . . . shows the critical values of T for various sizes of N. If the observed value of T is equal to lSidney Siegel, Nonparametric Statistics for the Behavioral Sciences (New York: McGraw—Hill Book Company, InC-, 1956), pp. 75-83. 2Ibid., p. 254. 1" \z 253 or less than that given in the table for a particular significance level on a parti— cular N, Ho (null hypothesis) may be re- jected at that level of significance. b. If N is larger than 25, compute the value of 2 as defined by the formula . . T- /N(N + l)(2N + 1) 2H "“‘ Determine the associated probability under Ho by referring to . . . (Table of Proba— bilities Associated with Values as Extreme as Observed Values of z in the Normal Distribution).3 For a two—tailed test, double the p shown. If the p thus obtained is equal to or lesi than a (significance level), reject Ho. N(N + l) ‘ —_——7T_.- Z: Mann—Whitney U Test5 The Mann-Whitney statistic is a nonparametric test to determine the probability that two independent samples, Which may be of unequal size, have come from identical populations. Procedure 1. Determine the value of n and n . n = the number of cases in the smaller group; n2 = the number of cases in the larger group. 2. Rank together the scores for both groups, assigning the rank of l to the score Wthh is algebraically lowest. Ranks range from 1 to N = n + n . Assign tied observations the average of the tied ranks. 3Ibid., p. 247. “Ibid., p. 83. 51bid., pp. 116-127. 1 y‘éfl / 1' 254 % Determine the value of U . . . (by the formulae n (n + l) _ l 1 U - nln2 + 2 — R1 or, equivalently, n (n + l) _ 2 2 U — nln2 + 2 — R2 The method for determining the significance of the observed value of U depends on the size of n . a. If n is 8 or less, the exact probability associated with a value as small as the observed value of U is shown in (the Table of Probabilities Associated gith Values of U in the Mann-Whitney Test). If (the) ob- served value of U is not shown in (the table), it is U1 and should be transformed to U by the formula b. If n2 is between 9 and 20, the significance of any observed value of U may be determined by reference to (the Table of Critical Values of U in the Mann—Whitney Test).7 If (the) observed value of U is larger than hing/2, it is U1; apply formula . . . (U = nln2 — U1) for a transformation. 0. If n2 is larger than 20, the probability associated with a value as extreme as the observed value of U may be determined by computing the value of z by the formula . . . n1H2 U 2 Z = /(nl)(n2)(nl + n2 + 1) l2 6Ibid., pp. 271—273. 7Ibid., pp. 274—277. 255 and testing this value by reference to (the Table of Probabilities Associated with Values as Extreme as Observe Values of z in the Normal Distribution). For a two— tailed test, double the p shown in that table. If the proportion of ties is very large or if the obtained p is very close to a, apply the correction for ties; i.e., use formula Z = , @sltas-w] -- If the observed value of U has an associated probability equal to or less than a, reject Ho in favor of H1' Ibid., p. 2M7. 9Ibid., p. 126. .4 ’* k NS “NW ” “11111111111if