A STUDY OF MARKET SEGMENTATION OF URBAN IN-HOME SHOPPERS BY SOCIOECONOMIC AND SOCIOPSYCHOLOGICAL CHARACTERISTICS Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY ISABELLA CLARA MANTOVANI CUNNINGHAM 1972 LIBRARY new , , Michigan State IIIIII/III!IIIII/III/II/I/I/II/IllIII/IIIIIII/I/II '2. ‘ - ? ”WW 3 1293 10415 6230 This is to certify that the thesis entitled A STUDY OF MARKET SEGMENTATION OF URBAN IN-HOME SHOPPERS BY SOCIOECONOMIC AND SOCIOPSYCHOLOGICAL CHARACTERISTICS presented by Isabella Clara Mantovani Cunningham has been accepted towards fulfillment of the requirements for Ph.D. degreein Marketing and Transportation Administration Date 4"9‘72 0-7639 § amour“; av 2‘: IIUAII 8: SONY “ mm mm m. - LIBRARY BINDERS SpeciaI Permission HQ renewal m» 313.1994; . ABSTRACT A STUDY OF MARKET SEGMENTATION OF URBAN IN-HOME SHOPPERS BY SOCIOECONOMIC AND SOCIOPSYCHOLOGICAL CHARACTERISTICS BY Isabella Clara Mantovani Cunningham Retailers in the United States have demonstrated their creativity by the introduction of innovative marketing techniques. Market segmentation is among the more pOpular marketing tools utilized by businessmen. The present study examines the delineation of market segments through measure- ment of socioeconomic and sociOpsychological variables. The research is limited to an investigation of individuals who shop at home utilizing mail and/or telephone. A disproportionate stratified sample of one thousand residents of Lansing and East Lansing, Michigan, was used in the study. Each individual in the sample was mailed a ques- tionnaire. Five hundred and nineteen responses were received and analyzed. Linear discriminant and canonical analysis were used to test the results of the study. Two sets of dependent variables were used in the research. The first set was used to test differences between peOple who shop at home and those who do not. The dependent variables used for this purpose were: Isabella Clara Mantovani Cunningham active in-home shoppers inactive in-home shoppers very active in-home shoppers very inactive in-home shOppers. The classification was based on the frequency of purchase by the respondents. The second set of dependent variables was used to verify the existence of market segments within the in-home shOpping market. The variables were: large catalogue and department stores club-of-the-month type of organizations novelty retailers specialty in-home retailers credit card organizations newspaper and magazine advertisements. The independent variables used in the research were the following: 1. ooqmuwnw 0 Social Class A. Total family income B. Occupation of the head of the household C. Level of education of the head of the household Family Life-Cycle A. Marital status of the homemaker B. Employment status of the homemaker C. Age of the head of the household D. Age and number of children living at home Adventuresomeness Cosmopolitanism Trust in People Tendency to Impulse Buying Attitude Toward Credit Conservatism. The thesis major findings are indicated below: 1. From the socioeconomic variables, and using discriminant analysis it was found that individuals who shop at home by mail and/or by telephone appeared to come from a higher social class than those individuals in the sample classified as not in-home shOppers. The Variable family Isabella Clara Mantovani Cunningham life-cycle was not useful in differentiating in-home shoppers from not in-home shOppers. 2. Linear discriminant analysis also showed that with respect to the sociOpsychological characteristics, the respondents classified as in-home shoppers appeared to be more cosmopolitan, less conservative, and had a more favorable attitude toward credit than did the not in-home shoppers. 3. The canonical analysis distinguished three market segments within the in-home shopping market. The first segment shops from large catalogue outlets, club-of- the-month outlets, and specialty outlets. These respondents come from a relatively high income, educational, occupa- tional, and social class level, and they display adventure- some attitudes. The second segment shOps at novelty outlets and does not shOp at specialty outlets. These individuals tend not to trust people, have local orientation, and are in the early stages of the family life-cycle. The third segment shops by mail and/or by telephone from specialty and credit card outlets. These respondents have the same characteristics as the respondents in the first segment, except they are less adventuresome. It is significant to note that some of the vari- ables that were of little use in the study for differen- tiating in-home shOppers from those respondents who did not shop at home, were useful in distinguishing submarket Isabella Clara Mantovani Cunningham segments, thus, a combination of variables is required to measure the degree of in-home shopping and potential market segments. The results of the present study cannot be general- ized beyond the sample and its related geographical location. However, it is concluded that, at least for the sample analyzed, market segmentation can be achieved by using a combination of socioeconomic and sociopsychological vari- ables. Perhaps future research will provide the practi- tioner and the marketing scholar with supporting data to generalization of the results. When this occurs, retailers in the in-home shOpping market will consider both socio- economic and sociOpsychological variables in making deci- sions about their marketing strategies. A STUDY OF MARKET SEGMENTATION OF URBAN IN-HOME SHOPPERS BY SOCIOECONOMIC AND SOCIOPSYCHOLOGICAL CHARACTERISTICS BY Isabella Clara Mantovani Cunningham 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 Administration 1972 r’ c Copyright by ISABELLA C. M. CUNNINGHAM 1972 ACKNOWLEDGMENTS The completion of a doctoral dissertation represents the combination of years of collective and individual efforts. The author's work alone is by no means the only factor responsible for this accomplishment. This offers, therefore, an Opportunity to acknowledge the author's recognition to the many individuals who were instrumental in the successful completion of the doctoral degree. Dr. Donal J. Bowersox, Professor of Marketing and Transportation Administration, Michigan State University, as chairman of the research committee and academic advisor contributed substantially with his close guidance and advice throughout the author's doctoral work. Dr. Bowersoxksexample, time and effort will always be remembered and sincerely appreciated. Dr. Donald A. Taylor, Professor and Chairman of the Department of Marketing and Transportation Administration, Michigan State University, has always given generous assis- tance to the author. As a member of the research committee Dr. Taylor was very helpful in the develOpment of the study. His assistance and advices have provided incentives to successfully accomplish the doctoral program. iii Dr. Stanley C. Hollander, Professor of Marketing and Transportation Administration, Michigan State University, as a member of the research committee, and as professor of the author in various doctoral courses has given continuous and valuable assistance to the study. Dr. Hollander's example of professional and scholar develOpment will always be a stimulus for the author's future career. Dr. James F. Rainey, Assistant Dean, College of Business and Associate Professor of Business Law, with his friendship and encouragement has significantly influenced and helped the author's work throughout the doctoral program. Dr. Rainey's help is greatly appreciated. Dr. William H. Cunningham, Assistant Professor of Marketing, University of Texas at Austin, my husband, has been a constant source of encouragement during all the doctoral program. His help and understanding has been of great value for the completion of the research. Without him, the program would not have been completed. To all the colleagues and friends, too numerous to individually recognize, who have generously offered their time, effort, and friendship to the author, which helped her in correcting her errors, and successfully accomplishing the tasks of this long and laborious journey to the doctoral degree, the author wishes to offer her sentiments of thanks and appreciation. iv TABLE OF CONTENTS ACKNOWLEDGMENTS . O O C C C O O O 0 O O C O O O 0 LIST OF TABLES O O O O O O O O O O O O O O O O 0 Chapter I. PROBLEM DELINEATION . . . . . . . . . . . A Brief Historical Perspective of In—Home Shopping . . . . . . . . . . . . . . Nature of the Problem . . . . . . . . . Purpose of the Study . . . . . . . Statement of the Problem . . . . . Hypotheses and Subhypotheses . . . Methodology . . . . . . . . . . . . Limitations of the Study . . . . . Contributions of the Study to Marketing Theory . . . . . . . . . . . . . Contributions of the Study to Marketing Practice . . . . . . . . . . . . . . Organization . . . . . . . . . . . . . II. LITERATURE REVIEW . . . . . . . . . . . . Mail Order and In-Home ShOpping . . . . The Origin and Development of Mail Order Sales . . . . . . . . . . . A Literature of "How to Make a Fortune" . . . . . . . . . . . . General In-Home Shopping Literature Social Class and Consumption Behavior . The Reputational Method . . . . . . The Sociometric Method . . . . . . The Occupational Method . . . . . . The Mu1tiple Index Method . . . . . The Application of Social Classes j to Marketing . . . . . . . . . . \ Family Life-Cycle and Consumption I Behavior . . . . . . . . . . . . . . Page iii viii ll l3 18 20 22 23 24 25 25 25 35 38 44 47 50 51 52 54 58 Chapter L/ Techniques of Classification of Family Life-Cycle . . . . . . . Consumption and Family Life-Cycle . Personality Variables . . . . . . . . Segmentation of the In-Home Shopping Market by Socioeconomic and Socio- psychological Variables . . . . III. r RESEARCH DESIGN 0 O O O O O O O O O O 0 Research Design Framework . . . . . . Identification of Variables . . . Independent Variables . . . . . . Dependent Variables . . . . . . . Sample Selection . . . . . . . . . . Data Analysis . . . . . . . . . . . . Linear Discriminant Analysis . . Canonical Analysis . . . . . . . Summary . . . . . . . . . . . . . . . IV. PRESENTATION OF FINDINGS . . . . . . . Active vs. Inactive In-Home Shoppers Socioeconomic and Demographic Variables . . . . . . . . . . . Sociopsychological Variables . . Linear Discriminant Analysis . . Very Active vs. Very Inactive In-Home Shoppers . . . . . . . . . . . . . Socioeconomic and Demographic Variables . . . . . . . . . . . SociOpsychological Variables . . Discriminant Analysis . . . . . . Canonical Analysis . . . . . . . . . Conclusion . . . . . . . . . . . . . V. EVALUATION OF THE RESEARCH HYPOTHESES . Evaluation of the Selection of the Independent Variables . . . . . . . Evaluation of the Research Hypotheses Evaluation of the Second Set of Hypotheses . . . . . . . . . . . . First Group . . . . . . . . . . . Second Group . . . . . . . . . . Third Group . . . . . . . . . . . vi Page 61 64 65 68 7O 7O 7O 71 74 77 8O 80 84 87 88 88 88 101 104 111 111 123 123 128 133 134 134 137 148 148 154 159 Chapter Analysis of the Research Findings in View of the Research Objectives . . . . . Summary . . . VI. SUMMARY AND CONCLUSIONS . . . . . . . . . . . Summary of the Research . . . . . . . . . . Conclusions from the Findings . . . . . . . Implications of the Theory . . Implications of the Management Suggested Areas for BIBLIOGRAPHY . . . . . Appendix A. Letter of Introduction B. Follow-Up Card C. Questionnaire . Research for Marketing Research for Marketing Further Research . . . vii Page 164 165 166 166 169 171 175 178 180 188 189 190 10. 11. 12. 13. 14. LIST OF TABLES Growth of Manufacturing in the United States from 1860 to 1900 . . . . . . . . . . . . . United States Population and Income-- 1879-1910 . . . . . . . . . . . . . . . . . Yearly Sales of Leading Mail Order concerns, 1912-1928 C O O O O O O O O O O O Socioeconomic Characteristics of Active and Inactive In-Home Shoppers . . . . . . . Demographic Characteristics of Active and Inactive In-Home Shoppers . . . . . . . . . Occupation of the Head of the Household of Active and Inactive In-Home Shoppers . . Income Levels of Active and Inactive In-Home ShOppers . . . . . . . . . . . . . . Education of the Head of the Household of Active and Inactive In-Home Shoppers . . . . Social Class of Active and Inactive In-Home Shoppers O O O O O O O O O O O O O O O O O 0 Employment of the Homemaker of Active and Inactive III-Home Shoppers o o o o o o o o 0 Marital Status of Active and Inactive In-Home Shoppers o o o o o o o o o o o o o c Number of Years Married of Active and Inactive In-Home Shoppers . . . . . . . . . Age of the Head of the Household of Active and Inactive In-Home Shoppers . . . . Family Life-Cycle of Active and Inactive In-Home Shoppers . . . . . . . . . . . . . . viii Page 28 30 34 89 90 92 93 94 96 97 98 99 100 102 Table 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. Page Sociopsychological Variables of Active and Inactive In-Home Shoppers . . . . . . . . . . 103 Correlation Coefficients with the Discrim- inant Score and F-Test for the Difference in Group Means for Active and Inactive In-Home Shoppers . . . . . . . . . . . . . . . 105 Discriminant Classification Matrix for Active and Inactive In-Home Shoppers . . . . . . . . 107 Wilks' Lambda Test and Chi Square Test for Statistically Significant Differences Achieved by the Four Independent Cross Validated Runs on the Active and Inactive In-Home Shoppers . . . . . . . . . . . . . . . 108 Correlation Coefficients with the Discrim- inant Score and F-Test for 60 Percent of the Active and Inactive In-Home Shoppers for Four Independent Cross Validated Samples . . . . . . . . . . . . . . . . . . . 110 Demographic Characteristics of Very Active In-Home Shoppers and Very Inactive In-Home Shoppers O O O O O I O O O O O O O I O O O O O 112 Socioeconomic Characteristics of Very Active In-Home Shoppers and Very Inactive In-Home Shoppers O O C O O O O O O O O O O O O O O O O 113 Occupation of the Head of the Household of Very Active and Very Inactive In—Home Shoppers O O O O O O O O O O O O O I O O 0 O O 114 Income Levels of Very Active and Very Inactive In-Home Shoppers . . . . . . . . . . 115 Education of the Head of the Household of Very Active and Very Inactive In—Home Shoppers O O O O O O O O O I O O O O O O O O O 116 Social Class of Very Active and Very Inactive III-”Home Shoppers o o o o o o o o o o o o o o o 117 Employment of the Homemaker of Very Active and Very Inactive In-Home Shoppers . . . . . . 118 ix Table 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. Marital Status of Very Active and Very Inactive In-Home ShOppers . . . . . . . . . Number of Years Married of Very Active and Very Inactive In-Home Shoppers . . . . . Age of the Head of the Household of Very Active and Very Inactive In-Home Shoppers . Family Life Cycle of Very Active and Very Inactive In-Home Shoppers . . . . . . . . . Discriminant Classification Matrix for Very Active and Very Inactive In-Home Shoppers . Correlation Coefficients with the Discrim- inant Score and F-Test for the Differences in Group Means for Very Active and Very Inactive In-Home Shoppers . . . . . . . . . Wilk's Lambda and Chi Square Test for Statistically Significant Differences Achieved by the Three Independent Cross Validated Runs on the Very Active and Very Inactive In-Home Shoppers . . . . . . . Correlation Coefficients with the Discrim- inant Score and F-Test for 60 Percent of the Very Active and Very Inactive In-Home Shoppers for Three Independent Cross Validated Random Samples . . . . . . . . . . Results of the Canonical Analysis for Active In-Home Shoppers . . . . . . . . . . Results of the Canonical Analysis for the Very Active In-Home Shoppers . . . . . . Page 119 120 121 122 124 125 126 127 129 132 CHAPTER I PROBLEM DELINEATION A Brief Historical Perspective of In-Home ShOpping When analyzing the structure of retailing in the western world, it is interesting to note the many different types of institutions that exist. These include: super- markets, department stores, catalogue stores, discount stores, and specialty stores, among others. The development of such varied types of retailing institutions has accompanied the evolution of the physical production of goods and services. These institutions have resulted from changes in consumption patterns. Retailing serves a continuing function as a link between producers 7‘s}: \ [y",.. I and consumers, aiding the removal of the discrepancies which invariably exist between them.1 The in—home shopping segment of retailing has expanded dramatically since the end of World War II.2 1A. C. R. Dreesman, "Patterns of Evolution in Retailing," Journal of Retailing, XLIV, No. 1 (Spring, 1968), 64-67. 2Jerome E. McCarthy, Basic Marketing (Revised ed.; Homewood, 111.: Richard D. Irwin, Inc., 1964), p. 499. In the ten-year period from 1961 to 1970, in-home shopping sales grew from 1.9 to 3.8 billion dollars.1 This near doubling of magnitude is particularly impressive in that aggregate retail sales during the same decade experienced a comparative increase of 66 percent.2 In addition, the future growth potential of in-home shopping has been singled out by several as one of the high potential areas for revo- lutionary changes in marketing practices during the remainder of the twentieth century.3 The development of the mail order business on a large scale during the 1870's was one of the most remarkable commercial events in the evolution of the existing retail structure. It was customary during American colonial times for wealthy individuals to order many types of goods from EurOpean countries. This practice continued until the Revolutionary War. After that war, impoverished conditions existent in the United States resulted in a considerable decrease in the overseas purchasing practices. 1U.S. Department of Commerce, "Mail Order Sales of Department Store Merchandise by Mail Order Companies or Mail Divisions," Survey of Current Business, LI, No. 9 (September 1970), 5-11. zIbid. 3"Tele-Purchasing--Major Trend in Retailing?" Forbes, October 15, 1967, pp. 56-59; and Alton F. Doody and William R. Davidson, "Next Revolution in Retailing," Harvard Business Review, May-June, 1967, pp. 4-20. In the early part of the nineteenth century, American retail trade developed and eXpanded rapidly. Retail stores sold primarily imported products, and later, as they became available, domestically manufactufed mer—_ chandise. Settlers in the western frontier areas frequently ordered goods by mail for transport over wagon routes and later for transport by canal and railroads from the retail stores of the large eastern centers. It appears that retail stores handled this early mail order entirely as a sideline, and as a convenience to former customers and others Who had moved to distant parts of the country.1 Paul H. Nystrom in The Economics of Retailing, states that the earliest intensive, wide-scale effort to secure business exclusively by mail was made by E. C. Allen of Augusta, Maine, who initiated national advertising of specialty products in 1870.2 His efforts were imitated when several mail order periodicals began publication. In 1872, Montgomery Ward established the first general merchandise mail order house in Chicago, followed by Sears and Roebuck Company and Spiegel Company. Mail order distribution continued to expand during the 1880's and 1890's. Department stores attempted to enter 1Paul H. Nystrom, The Economics of Retailing (3rd. ed.: New York: The Ronald Press C0,, 1930), Vol. I, Ch. III, pp. 58-680 2Ibid., Vol. I, p. 175. this business by establishing mail order departments. Until approximately the beginning of World War II, it was the customary practice of these department stores to print customer mail order catalogues. In this same period, specialty houses began to utilize mail order techniques to distribute their products. Among the initial commodities sold by mail were books, magazines, garden products, cosmetics, and novelties.l The retailing literature shows a general consensus concerning the major factors which contributed to the early growth of mail order retailing. These factors include: (1) the existence of a high literacy rate, (2) the ineffi- ciency of some local merchants, (3) the increased earning power of the people, and (4) the growth of industrialization leading to the intensive distribution of goods.2 In the early 1900's the nation experienced a signif- icant migration of people from rural to urban areas. Urban concentration provided the Opportunity for the development of department stores which would compete directly with the mail order houses. As the use of the automobile increased, downtown stores became more accessible to urban and suburban shoppers alike. At the same time, the telephone proved its 1Ibid., Vol. I, pp. 180-205. 2See: Ibid., Vol. I; Paul Tuey, Elements of Retail Salesmanship (New York: The McMillan Co., 1925); and Boris Emmet and John Jeuck, Catalogues and Counters (Chicago: The University of Chicago Press, 1950). .2“. utility as a convenient substitute for in-home shopping by mail. In spite of the dramatic social and economic changes that have occurred in the United States since the beginning Of the twentieth century, the volume Of in-home shopping has not declined. Rather, catalogue mail order houses have experienced continued growth in their in-home sales. In view of the very large migration from rural areas, it seems logical to conclude that this increase in sales must be attributable to customers living in urban and suburban areas. As a consequence, some large mail order houses have located many Of their new catalogue stores in small towns, and the suburban regions of metropolitan areas. Increasing consumer affluence has created new mar- kets for products that can be sold competitively by direct mail firms. The use Of computers has simplified in-home retailing Operations and has decreased the time required to complete a mail order sale. The adoption Of credit cards has simplified activities of ordering and billing by mail. In the last few years, several in-home sales orga- nizations have modified their system to Offer the consumer more convenience in in-home shopping. Some firms have adOpted computerized order forms to decrease the lead time between ordering and delivery. Other improvements have been the use Of credit cards for billing purposes, and special services such as gift wrapping, and others. All of these variables have contributed to the important position of the in-home shopping market within the retail activities in the United States. Nature of the Problem Trade practices in America have undergone substan- tial changes since the days of the first European settle- ments. Steven C. Brandtl has described the evolution of the United States distribution system as a three step process. First, the "Open market" developed where producers were Operating within economic and geographical constraints that dictated the use of "natural" segments as a means for allocating their efforts. Second, the "mass market" evolved where producers could reach larger segments of the pOpula- tion because of improvements in transportation and increased levels of disposable income of consumers. More sophisticated segmentation was then employed, based on such variables as age, family size and ethnic background. Third, the "frag- mented market" developed where the mass market disappears and a more fragmented market starts to grow, along with a need for more sophisticated methods of segmentation, such as the use of sociopsychological variables.2 1Steven C. Brandt, "Dissecting the Segmentation Syndrome," Journal of Marketing, XXX (October, 1966), 22. 21bid. Since the advent of mass production, early in the 1900’s, the strategy of market segmentation has been a managerial practice. Understanding that a market is com- posed Of a set of individuals, none of which have identical desires, needs or motives is today common knowledge among businessmen. The desire of bringing together heterogeneous markets and the economic requirements of standardized pro- duction Of goods by manufacturers has caused the development of strategies with the main objective of grouping sub mar- kets with more or less homogeneous characteristics. This, in turn, permits manufacturers to design their products for relatively distinct market segments. Empirical Observation alone is sufficient to notice the fruits of market segmentation as it is practiced by major national manufacturers of consumer goods. For example, the cigarette market is substantially changed since World War II. Previous to the war, four brands of cigarettes dominated the market, all of which were packed without filters and non-mentholated. The enormous variety of ciga- rettes now existing in the same market shows the large extent to which the market has been segmented by the manufacturers. Management's widespread adoption of market segmen- tation--made Obvious by the postwar boom of new brands and sales outlets-—must be recognized as a marketing strategy tfliat takes account Of more than just simple economic and demographic changes. Motivations and attitudes, as well as personality characteristics have been used as criteria for segmenting markets.1 The heterogeneity in age, sex, income, social class, personality characteristics, motivation, attitudes, needs and desires exists among in—home shoppers as well as all other consumers of the American market. With this under- standing, it seems natural to assume that it would be pos- sible to segment the in-home shoppers into smaller and homogeneous groups. In order to accomplish such a task, variables such as socioeconomic characteristics, education, income level, as well as sociopsychological traits should be Observed, considered and analyzed. Market segmentation is considered a valuable tool by businessmen and scholars. The in-home shopping market in the United States is large and has no geographical boundaries. These factors suggest that a detailed inves- tigation Of applications of market segmentation in the in- home shopping market may lead to an improved match of market Offerings to various in-home customers market segments. Purpose of the Study Within the framework of the foregoing discussion, the purpose Of the prOposed study is to analyze the lPhilip Kotler, Marketing Management--Analysis, Planning and Control (Englewood Cliffs, N.J.: Prentice- existence Of segments in the in-home shOpping market. The study will also attempt to describe the characteristics of these segments taking into account not only socioeconomic and demographic variables, but also sociopsychological variables.l According to Wendell Smith, "segmentation is based upon developments on the demand side of the market and represents a rational and more precise adjustment of product and marketing effort to consumer or user requirements."2 The idea that markets can be profitably segmented has received widespread acceptance. The problem confronted by marketers is that Of deciding which of the virtually limit- less alternatives is likely tO be the most productive Kotler suggests several bases for segmenting buyers.3 (1) socioeconomic variables such as age, sex, family size, income, occupation, family life-cycle, religion, race, nationality and social class; (2) geographical variables, such as region, county size, city size, density and climate; (3) personality variables such as compulsiveness, gregar- iousness, autonomy, conservatism, authoritarianism, leader- ship, ambitiousness, etc., (4) buyer behavior variables 1Ed Burnett, "How to Select Lists that Reach People with Ability to Buy," The Reporter of Direct Mail Advertising, April, 1969, p. 34. 2Wendell R. Smith, "Product Differentiation and Market Segmentation as Alternative Product Strategies," Journal of Marketing, XXI (July, 1956), 3-8. 3Kotler, op. cit., pp. 47—53. 10 which describe aspects of the buyers' relation to a specific product. Examples of buyer behavior variables are: usage rate, buyer motive, brand loyalty, buyer class, end use, channel loyalty, price sensitivity, advertising sensitivity, etc. Russell I. Haley describes a new approach to market "1 Benefit segmentation segmentation: "benefit segmentation. is based on the assumption that the benefits that peOple are seeking in consuming a given product are the basic reasons for the existence of market segments. An application Of combined segmentation criteria should provide more homogeneity within segments than if only socioeconomic variables were utilized. Psychological and attitudinal variables may be useful tools in discriminating among subjects classified in the same age group, the same social class, and the same educational level. Although the most recent developments in the study and use of mailing lists have shown a tendency toward segmentation strategy, no reference has been found as to the use of psychological and attitudinal variables in delineating the in-home shopping market.2 The success achieved through the utilization of psychological variables to determine and define submarkets for consumer goods suggests that motivational, attitudinal lRussel I. Haley, "Benefit Segmentation: A Decision- Oriented Research Tool," Journal of Marketing, XXXII (July, 1968), 30-35. 2Burnett, op. cit., p. 34. 11 and personality characteristics may prove to be of considerable importance in segmenting in-home shoppers. Statement of the Problem The in—home shopping market consists of those individuals who purchase merchandise without going to a retail outlet. This definition encompasses catalogue pur- chasers, mail order purchasers, telephone purchasers, and all types of direct mail purchasers. More specifically, in-home shoppers include all those individuals who buy without making any physical inspection of the goods, or without any face to face contact with a salesman or a rep- resentative of the firm which is selling the product. It is essential to this definition that the customer not have any contact with the goods before the actual purchase; the consumer's decision must be based primarily on a written description, drawing and/or photograph of the product. Once the purchase is made, it is irrelevant whether the goods are delivered to the customer, or whether he must pick them up at the local store or general delivery outlet. Excluded from this definition of in-home shoppers are those who purchase products from door-to-door salesmen. While the usefulness of segmentation strategies in marketing is generally recognized, little research has been devoted to characterizing the in-home shOpper as a distinct group among the American consumers. Also, little research 12 has been made to verify the existence of segments within the in-home shOpping market. The objectives of this research are, therefore, the following: 1. To determine whether significant differences exist between individuals who shop at home by telephone and/or mail, and those who do not. 2. To verify the existence of market segments within the in-home shOpping market. 3. To suggest a method for establishing segmentation criteria for firms in the in-home shopping market in order to facilitate their selling strategies. The independent variables chosen differences between in-home shoppers and shop at home, as well as to indicate the ments within the in-home shopping market family life-cycle, attitude toward risk, to verify potential peOple who do not existence of seg- are: social class, adventuresomeness, cosmopolitanism, trust in people, conservatism, attitude toward credit and tendency for impulse buying. Social class is selected as an independent variable because previous market segmentation studies have shown its usefulness as an instrument of differentiation between mar- ket segments. One of the most well known studies utilizing social class to segment consumers was the Chicago study by Martineau.l In the same manner, family life-cycle has been 1Pierre Martineau, "Social Class and Spending Behav— ior," Journal of Marketing, XXIII (October, 1958), 123-125. 13 used in the past as a predictor variable of consumption 1 For this reason, they are selected as an behavior. independent variable for the present research. The sociopsychological variables: attitude toward risk, adventuresomeness, cosmopolitanism, trust in people, conservatism, attitude“tOward credit and tendency for impulse buying are selected inferentially by the researcher. To the researcher's knowledge, these variables have not been previously used in in-home shopping studies. Four of these variables: adventuresomeness, conservatism, trust in peo- ple, and cosmOpOlitanism have been utilized in sociological studies pertaining to behavioral patterns. It is felt that the sociOpsychological variables might be able to discrim- inate between in-home shoppers and people who do not shop at home, as well as indicate the existence Of segments within the in-home shopping market. / ‘ Hypotheses and Subhypotheses : .1 y The research tests the validity of two major sets of hypotheses. The first set of hypotheses refers to the existence of differences between in-home shoppers and those individuals who do not shOp at home. The second set Of hypotheses refers to the existence of distinct submarket 1Lincoln Clark, ed., Consumer Behavior II: The Life Cycle and Consumer Behavior (New York: New York University Press, 1955). 14 segments within the total in-home shopping market. Both sets Of hypotheses are written in the null form in order to facilitate testing. I. In-Home Shoppers vs. All Other Consumers 1. Social Class: Ceteris paribus, among all respon- dents there is no significant difference between peOple who shop at home and people who do not shop at home, when classified by their respective social classes. A. Occupation: Ceteris paribus, among all respon- dents, in-home shOppers cannot be differentiated from peOple who do not shop at home, when clas- sified by their respective occupation. B. Income Level: Ceteris paribus, among all respondents, there is no significant difference between in-home shoppers and people who do not shop at home, when classified by their respec- tive income level. C. Educational Level: Ceteris paribus, among all respondents there is no significant difference between in-home shoppers and peOple who do not shop at home, when classified by their respec- tive educational level. 2. Family Life Cycle: Ceteris paribus, among all respondents there is no significant difference 3. 15 between peOple who shop at home and people who do not, when classified by their stage in the family cycle. A. Marital Status: Ceteris paribus, among all respondents, there is no significant difference between in-home shoppers and those who do not shop at home, when classified by their reSpec- tive marital status. B. Age of the Household Head: Ceteris paribus, among all respondents, there is no significant difference between in-home shoppers and people who do not shop at home, when classified by the age of the household head. C. Employment Status of the Homemaker: Ceteris paribus, among all respondents there is no significant difference between in-home shoppers and people who do not shop at home, when classi- fied by the employment status of the homemaker. D. Number of Years Married: Ceteris paribus, among all respondents, there is no significant differ- ences between in-home shOppers and people who do not shop at home, when classified by the number Of years they were married. Adventuresomeness: Ceteris paribus, among all respondents there is no significant difference between in-home shoppers and peOple who do not 16 shOp at home, when classified by their respective adventuresomeness. Cosm0politanism: Ceteris paribus, among all respondents there is no significant difference between people who shop at home and people who do not shOp at home, when classified by their respec- tive degree of cosmOpOlitanism. Trust in People: Ceteris paribus, among all respondents there is no significant difference between people who shOp at home and peOple who do not shop at home, when classified by their respec- tive trust in peOple. Conservatism: Ceteris paribus, among all respon- dents there is no significant difference between people who shop at home and people who do not shop at home, when classified by their respective con- servative-liberal attitudes. Attitude Toward Credit: Ceteris paribus, among all respondents, there is no significant differences between people who shop at home and people who do not shOp at home, when classified by their respec- tive attitude toward credit. Tendency for Impulse Buying: Ceteris paribus, among all respondents, there is no significant difference between peOple who shOp at home and people who do not shOp at home, when classified by their respec- tive tendency for impulse buying. 17 II. Comparison Between Segments of In—Home Shoppers. 1. Social Class: Ceteris paribus, among all in-home shoppers, there is no significant difference when classified by their respective social class. A. Occupation: Ceteris paribus, among all in—home shOppers there is no significant difference when classified by their respective occupation. B. Income Level: Ceteris paribus, among all in- home shoppers, there is no significant differ- ence when classified by their respective income level. C. Educational Level: Ceteris paribus, among all in-home shoppers, there is no significant dif- ferences, when classified by their respective educational level. Family Life Cycle: Ceteris paribus, among all in- home shoppers, there is no significant difference, when classified by their stage in the family cycle Adventuresomeness: Ceteris paribus, among all in- home shoppers, there is no significant difference when classified by their respective adventuresome- ness. Cosmopolitanism: Ceteris paribus, among all in-home shoppers, there is no significant difference when classified by their respective cosmopolitan attitudes. 18 5. Trust in People: Ceteris paribus, among all in—home shoppers there is no significant difference when classified by their respective trust in people. 6. Conservatism: Ceteris paribus, among all in-home shoppers there is no significant difference when classified by their respective conservative-liberal attitudes. 7. Attitude Toward Credit: Ceteris paribus, among all in-home shoppers there is no significant difference when classified by their respective attitude toward credit. 8. Tendency for Impulse Buying: Ceteris paribus, among all in-home shOppers, there is no significant dif- ference when classified by their respective tendency for impulse buying. Methodology The study was made in the State of Michigan. Data were collected in June and July, 1970, in the cities of Lansing and East Lansing. The location was chosen because it was felt that Lansing and East Lansing represent a rep- resentative spectrum of American society and, therefore, the results Of the research could be generalized on a limited basis to a broader pOpulation than merely the research area. The Lansing and East Lansing area was divided into five income tracts. Two hundred families were randomly selected from each tract to participate in the research. 19 The data were first tested with linear discriminant analysis to determine whether there were significant differ- ences between subjects who shopped at home and those who did not. The linear discriminant analysis test was applied in two steps. First active in-home shOppers were compared to inactive in-home shOppers. For purposes of the analysis defined as active in—home shOppers were all individuals who indicated in question 7, Section I of the questionnaire that they either purchased goods by mail or telephone "regularly" or "occasionally" in the previous two years. Inactive in- home shoppers were classified as being the individuals who indicated on question 7, Section I of the questionnaire that they "very rarely" or "never" bought goods by mail or by telephone during the previous two years. The second step of the linear discriminant analysis was to compare the very active and the very inactive in-home shOppers. For purposes Of the analysis defined as very active in-home shoppers were those subjects who indicated on question 7, Section I of the questionnaire that they "regularly" purchased goods by mail or telephone during the previous two years. Very inactive in-home shoppers were classified as the subjects who indicated in question 7, Section I Of the questionnaire that they "never" purchased goods by mail or telephone during the previous years. Canonical analysis was used to examine whether market segments exist within the in-home shOpping market. 20 The subjects were asked to indicate from which of the following in-home retailers they had purchased merchandise in the past by mail or by telephone: (1) large catalogue stores, (2) specialty mail order houses, (3) club-Of-the- month type of organization, (4) novelty mail order houses, (5) credit card type of organization, and (6) newspaper and magazine advertisements. The active in-home shoppers were analyzed with canonical analysis to determine if there were strong relationships among the subjects who indicated that they purchased goods from the same six types of retail out- lets. That is, if it could be demonstrated that the respon- dents who preferred the same in-home shopping outlets were strongly correlated with each other by similar characteris- tics, then they would probably represent a market segment. The very active in-home shOppers were also examined through canonical analysis, in order to verify the existence of market segments among them. Both discriminant analysis and canonical analysis are discussed in further detail in Chapter III. Limitations of the Study Generalizations derived from the proposed research are limited for the following reasons: 1. The data utilized in the research were collected at one point in time, instead of over a period of time. This means that consistency of shopping 21 habits of the sample is assumed, which may not have been correct if the data were collected in a longitudinal manner. The independent variables selected as predictors of in-home shOppers behavioral patterns were chosen following an inferencial criterion, rather than an experimental criterion. Accordingly, there may be several other demographic, attitudinal, motivational, and sociOpsychological variables which would be bet- ter predictors of in-home shOpping behavior, and consequently were not examined in the research. The non-inclusive list of variables, along with the pur- chasing patterns Of individuals in the United States, implies that careful consideration must be made before extrapolating the findings Of the research across geographical regions or to specific markets. The sample studied was 1,000 families and was con- fined tO urban areas Of 50,000 and 150,000 people, approximately. Because of this reason, and the fact that there are considerable differences Of shOpping habits among people living in different urban and suburban shopping areas, the accuracy with which regional purchasing patterns may be drawn from the findings Of the study are limited to the degree to which discrepancies exist between the normalized values Obtained through the sample and local char- acteristics. 22 Contributions of the Study to Marketing Theogy The contribution of the research to marketing theory rests on its contribution to the description of purchasing patterns Of consumers, within the emerging body of marketing thought. Contributions to theoretical development may be classified into three major types: generation of new hypotheses and theories, prediction of future occurrences and description of existing phenomena. Therefore, an addi- tion to the description of consumer behavior patterns, as to what concerns the in-home shOppers is a first contribu- tion to the theoretical body of knowledge. Secondly, a detailed description constitutes the first prerequisite for future predictions of variations in the behavioral patterns of in-home shoppers. The foundations for such prediction are a possible outcome of the present research. Lastly, a more important contribution is the characterization Of in- home shOppers in terms of social class, family life-cycle and the sociOpsychological variables: attitude toward risk, adventuresomeness, tendency toward impulse buying, attitude toward credit, cosmOpOlitanism, conservatism and trust in people, which is vital for establishing the background from which new theories of a particular form of consumer behavior might be generated. Accordingly, the characterization Of in-home shoppers according to socioeconomic and sociOpsy- chological Variables is a necessary prerequisite to the 23 generation of testable hypotheses concerning determinants of their purchasing patterns and behavior. Contributions Ofythe Study to Marketing Practice It is expected that the research will reveal differences between people who shop at home and those who do not; and will locate relatively distinct submarkets within the in-home shopping market. It is hypothesized that the psychological scales will play a significant role in differentiating these markets. Therefore, the first con- tribution Of the study will be to indicate the feasibility of utilizing sociopsychological variables to segment the in-home shopping market composed Of those peOple who pur- chase goods at home by mail and/or by telephone. The second contribution Of the study is to increase the existing knowledge of consumer purchasing patterns. The research attempts to describe the in-home buyers' behavior. It also attempts to predict the reactions of future in-home buyers to provide a greater adaptability to their needs and wants by the retailer, through segmentation. A third contribution would be to try to add reasons why this modality of retailing should be given greater attention by middlemen interested in future develOpments of the United States market. The research will attempt to determine Opportunities for market entry by delineating the specific characteristics of market segments. 24 Finally, the results of the research might help marketing practitioners in selecting their marketing policies. The allocation of the marketing effort depends on the relative Opportunities existing in the market. Therefore, a more detailed description of consumers‘ atti- tudes and motivations, as well as their social class and family life-cycle, will provide a basis for a more appro— priate determination of prices, promotion and distribution by in-home sellers. Organization The remainder of the study is divided into five chapters. Chapter II presents a review of literature relevant to the research. In Chapter III the research design used to analyze the data Of the study is described. Chapter IV is a description of the research findings. Chapter V relates the research findings to their respective hypotheses. Chapter VI contains the summary and conclusions of the study, as well as suggestions for future research. CHAPTER II LITERATURE REVIEW The Objective of Chapter II is to provide a theoretical background for the research. The chapter is divided into five sections: Mail Order and In-Home ShOpping, Social Class and Consumption Behavior, Stage in the Family Life Cycle and Consumption Behavior, Personality Variables, and Segmentation of the In-Home Shopping Market by Socio- economic and Sociopsychological Variables. Mail Order and In-Home Shopping -1, ‘ L‘ The Origin and Development of Mail Order Sales The origin and develOpment of the first mail order and catalogue organizations at the end of the nineteenth century was a social and economic phenomenon which has been largely accounted for by historians and scholars. Because Of the importance which institutions such as Sears and Roe- buck, and Montgomery Ward have had in the retailing evolu- tion of the United States, special attention is given in this part Of the study to the development Of mail order and catalogue sales. An attempt is also made to situate the mail order houses within the total evolution of retailing. 25 26 ,’/“ Paul H. Nystrom describes the path of develOpment of retailing in America and the various factors which accompanied the appearance and development of the mail . order houses.l He defines five separate stages in the development of retail trade in America: . Prehistoric Indian Trade The trading post period The general merchandise store era The period of rise and development of single-line independent merchandise stores 5. The modern period of growth of large-scale retailing involving extensive capital in such institutions as department stores, mail order houses, and chain stores.2 thWNH o The nineteenth century witnessed the appearance of large-scale retailing institutions. Among them, the first mail order houses. Nystrom explains how this new type of distribution began: Shortly after the middle of the nineteenth century there were simultaneous beginning of several new types of retail institutions which in our time have come to occupy a very impor- tant place in the retail distribution of goods. Among these were the department stores, the mail order houses, and the chain stores.3 The large scale methods of retailing constituted a revolutionary change in the type of distribution system which existed in the United States prior to the nineteenth 1Paul H. Nystrom, The Economics of Retailing (3rd ed.; New York: The Ronald Press Company, 1930), Vol. I, Ch. IV. 21bid., Vol. I, p. 70. 3rbi45, Vol. I, pp. 84-86. 27 century. A careful examination of some social and economic changes which took place during the same period might pro- vide a tentative explanation of the reasons why the first mail catalogue stores were successful in their operation. One of the most important factors which contributed to the flourishing of new retail institutions was the eco- nomic prosperity of the United States during the years of 1860 to 1920, approximately. Some of the symptoms of our nation's economic growth are the improvements achieved in the transportation, agriculture and manufacturing industries. During the second half of the nineteenth century, transportation was dramatically improved in the United States. The decade of 1860-70 was the most prosperous period ever known for the railroads.1 Also, a new agri- cultural era started, the United States farming products began to be exported, especially cereals, increasing the economic opportunities of people engaged in the farming activity. The most significant feature of the economy during that period, however, was the rise of manufacturing. Its growth is better described by Table 1. One of the immediate consequences of the combined effects of these three economic occurrences was the fact that products of equal or better quality to those imported 11bid., Vol. I, pp. 84-86. 28 Table 1 GROWTH OF MANUFACTURING IN THE UNITED STATES FROM 1860 TO 1900a J Capital Invested Wage-Earners Value of Products Census Years (dollars) Employed (dollars) 1860 1,009,855,715 1,311,246 1,885,861,676 1870 2,188,208,769 2,053,996 4,232,325,442 1880 2,790,272,606 2,732,595 5,369,579,191 1890 6,525,156,486 4,251,613 9,372,437,283 1900 9,817,434,799 5,308,406 l3,004,400,l43 aRalph K. Wadsworth, Handbook of Mail Order Selling and Merchandising (Chicago: The Dartnell Corporation, 1918), p. 23. from European countries were at that time produced inter- nally, and made available to the great majority of the American population. Because of their greater income, and also thanks to the existence of transportation facilities, the farmers of the midwest could consider the purchase of commodities previously available only to the high, prosper- ous class. As industrial technology grew and was perfected, consumer goods began to be manufactured in greater quan- tities. Creative marketing innovations emerged as a consequence of more intense competition, and the producer's market was substituted by a new, more dynamic, consumer market.1 lWilliam Greenlaf, American Economic Growth Since 1860 (Columbia, S.C.: University of South Carolina Press, 1968), p. 265. 29 A second important factor for the development of the mail order retailing in the United States is the concentra- tion of population mainly in rural areas. The farmers' communities were also wealthy insofar as they represented a large proportion of the United States national income. Table 2 will better illustrate the distribution of popula- tion and income in the United States during the period from 1879 to 1910. From Table 2 we can see that the bulk of population, as well as a great portion of American wealth was still concentrated until 1910, in small rural communities. The American farmer, therefore, in spite of having considerable purchasing power, was limited as to the product that he could buy locally, unless he had the willingness and the physical means to travel to the central city to shop for his goods. The mail order business closed this spatial gap. Distant rural markets were reached with goods not normally carried by the general merchandise stores by utilizing railroads to transport the products for local delivery by the postal service.1 The early mail order literature is a documentary of the legendary ways by which the first mail order busi- nesses started their trade. Most authors agree that the 1Paul H. Nystrom, op. cit., Vol. I, p. 174. UNITED STATES POPULATION AND INCOME--1879-l910a 30 Table 2 Population Agriculture National Income Income Year Rural Urban (000,000) (000,000) 1879 --- --- 1,371 7,227 1880 36,026,048 14,129,735 --- --- 1889 --- --- 1,517 10,701 1890 40,841,449 22,106,265 --- --- 1899 --- --- 2,933 15,364 1900 45,834,654 30,159,921 3,034 16,158 1905 --- --- 3,678 21,428 1906 --- --- 4,029 23,165 1907 --— --- 4,214 24,403 1908 --- --- 4,621 23,458 1909 --- --- 5,311 28,700 1910 49,973,334 41,998,932 --- --- aBoris Emmet and John E. Jeuck, Catalogues and Counters (Chicago: University of Chicago Press, 1950), p. 12. 31 original idea was first develOped in the United States by E. C. Allen. The earliest intensive wide—scale efforts to secure business by mail seems to have been made by E. C. Allen of Augusta, Maine, who started national advertising on specialties such as recipes for washing powder, engravings, chro- mos and novelties, as early as 1870. Allen's success led to rapid imitation close at hand. Augusta, Maine, became for many years and con- tinued to be the most important mail order center in the United States. Several so-called "mail order periodicals" were established, such as the household magazine, Comfort, and a list of periodicals published by Vickery and Hill, carrying mail order advertising, and sent to patrons of mail order concerns all over the country. During the 1870's, a rapid growth of mail order selling paralleled a similar growth of canvassing and peddling activity that grew up during the same time.1 Montgomery Ward was very successful in his trade. Having had previous retailing experience, he decided to start an independent mail order business in Chicago in 1872. Ward took advantage of the "Granger" movement which started in the United States approximately at the same time. Its basic objective was to improve the economic conditions of the agricultural classes. Soon after he [Montgomery Ward] reached Chicago, the "Granger" movement began to spread rapidly throughout the Middle West. The chief purpose of the Order of the Patrons of Husbandry, which had but recently come into existence, was to improve economic conditions of the agricul- tural classes. Among the methods urged were consumers' c00perative stores and the distri- bution of goods directly from producers to consumers. The attention called to the farmer's problems by the Spread of this new movement, }Nystrom, op. cit., p. 176. 32 linked with his experience in farm trade while in St. Joseph, probably aided Montgomery Ward in perfecting his plans for mail order retailing. The first storeroom occupied by Ward in 1872 was a hayloft over a stable on East Kinzie Street. His entire capital was about $2,000, partly borrowed. A one-page handbill naming of various items of good and their prices was sent out by mail. In the years from 1872 to 1874, Ward issued ten of such one-page handbills. The first catalog, a small pamphlet of only eight pages, size 3" by 5", was issued in 1874. By 1876 the catalog was increased in size to 150 pages and pictures if the merchandise were issued for the first time. The sales for that year amounted to $300,000. By 1880, the sales had increased to $410,000. Ten years later, in 1890, the sales amounted to $2,411,000 and in 1900 to $8,886,000. By 1906, its sales volume had reached $18,000,000 and by 1913 it had risen to $40,000,000.1 The beginning of Mr. Sears business was also more a result of an opportunity which was brought up by chance than a long term detailed business plan. Nystrom tells the circumstances that made Mr. Richard W. Sears the founder of one of the most successful retail businesses of the United States. In 1884, he received a package from a mail order concern to be delivered C.O.D. to a cus- tomer in this little Minnesota town. (North Redwood, Minnesota, where Richard W. Sears was employed as a railroad station agent.) The customer could not be found and Mr. Sears notified the company. Their reply was to urge him to sell the watch to someone else at a profit of $2, and he did so. This venture led him to order more watches and sell them at a similar profit. The prospects of this business proved so attractive that in 1886 he left the employ of the railroad company and started a mail order watch and jewelry concern in Minneapolis. Early in 1887, he employed A. C. Roebuck as watch ‘Ibid., p. 178. 33 adjuster and in 1889 he made Roebuck his partner. Somewhat later this concern was moved to Chicago, and in 1890, when Mr. Sears was twenty seven years of age, he sold this concern for $100,000. As a condition of the sale he agreed not to establish another mail order concern in the city of Chicago under his own name for five years.1 However, Sears went back to Minneapolis and started another organization, which, after several name and location changes, became the gigantic retail structure that is at this time, known internationally.2 Nystrom proceeds in the description of the birth and development of several other mail order organizations. It is interesting to note the growth that the three major retail firms of that time reached in a few years. This can be viewed summarized in Table 3. The first literature which appeared about mail order businesses was primarily concerned in relating the history of the most prOSperous mail order retailers. Several fac- tors were listed as being favorable to the development of mail order selling. None of the articles and books examined explored theoretical or institutional approaches to the in- home shopping phenomenon of the early 1900's. ’Ibid., pp. 180-182. 2Ibid., p. 182. 34 Table 3 YEARLY SALES OF LEADING MAIL ORDER CONCERNS, 1912-1928a Sears, Roebuck Montgomery Ward National Bellas Year & Co. & Co. Hess & Co. 1912 77,116,859 35,000,000 10,911,866 1913 91,357,276 39,725,713 13,276,259 1914 96,024,754 41,042,486 15,164,727 1915 106,228,421 49,308,587 17,371,650 1916 137,200,802 62,044,336 21,554,230 1917 165,807,600 73,512,645 27,649,537 1918 181,655,829 76,166,848 33,485,015 1919 233,982,584 99,336,053 39,449,985 1920 233,856,872 101,745,270 47,704,428 1921 159,034,518 68,523,244 37,481,210 1922 160,648,152 84,738,826 45,357,566 1923 191,324,146 123,702,043 52,399,783 1924 199,545,862 150,045,065 49,225,804 1925 234,421,930 170,592,642 46,685,376 1926 248,550,341 183,800,865 42,872,399 1927 268,731,794 186,683,340 44,665,419 1928 319,773,787 214,350,446 44,649,102 aPaul H. Nystrom, The Economics of'RetaiZing (3rd ed.; New York: The Ronald Press Company, 1930), Vol. I, p. 179. 35 I A Literature of "How to Make a Fortune" As soon as the extent of the success of firms like Sears and Montgomery Ward became apparent, a number of publications began to emphasize the advantages of owning and managing a mail order concern. Some of those stressed the fact that "almost no capital" was required in order to sell by mail; others were enthusiastic about the "newness" of the idea in itself. Several pages covered with advices and information about prices, costs, railroad fees and schedules, how to advertise, how to write letters, how to use mail as a substitute for salesmen, types of products which could be sold by mail, and many other typical instruc- tions characterized the "how to" literature.1 Ralph Bartolomew gives a good example of the type of message this literature was trying to transmit when he calls the mail order business: "the hidden empire."2 Some of the appeal used concentrated on the idea of eliminating the middlemen profits and offerings, therefore offering products at a more reasonable price than those sold by the local retailers. A clear example of this approach can be seen in the following: lRalph Bartolomew, A Short Course in Direct Mail (New York: Direct Mail Division, Publishers Printing Company, 1928), pp. 62-63. 2Ibid., p. 54. 36 Deeply implanted in the instincts and thoughts of a large number of persons is the desire to buy goods at the source of supply, or as near to it as it is possible to get. The phrase "shop at the factory" gathers up what a large number of members of the public understand by this method of commerce.1 Another reason was added to strengthen the idea: the fact that, as mail order houses were able to reach a larger market because their location did not limit their ability to sell at a greater distance, they would sell larger quantities, have faster turnover of products and, consequently, would be able to maintain their costs at a lower level than the local general merchandise or specialty store.2 Attractive characteristics of the mail order busi- ness were the low capital requirements, the fact that it could be started on a very small scale, and that it could be commenced as a spare time occupation. However, it was claimed that a person "had to be fitted for the business" by temperament, by habit and by nature. A very vague defini- tion of these prerequisites is generally given by the lit- erature. For some authors, a little business eXperience was a very valuable asset, as we can see by the following state- ment: 1Albert E. Bull, Mail Order and Instalment Trading (London: Sir Isaac Pitman and Sons, 1926), Ch. I, p. 2. 2W. M. Leonard Berkwitz, The Encyclopedia of Mail Order Business (New York: Leonard Berkwitz, 1900), Ch. II. 37 . . . Any person who thinks of entering business on his own account, whether in this or any line, should endeavor to get some experience in trading in another way, before embarking on the undertak- ing or speculating much capital.1 Some of the counseling and advice in this portion of the literature are interesting and colorful. . . . Be a merchant before you are a manufacturer, is a wise advice. It is always well not only to find a market before you make for it, but also to be sure that the demand obtained is regular and will be likely continued.2 During the early 1900's many small mail order businesses on a home scale appeared in the United States. Those businesses used mail advertising or newspaper space to sell their products. "Postal bargains" made daily appearance in the newspapers. In order to avoid suspicion by the potential clients, "money back" guarantees were added to the offers.3 The proliferation of "how to" literature is a testimony of the popularity that the mail order trade had at the beginning of the twentieth century. A variation of this approach is the elaboration of textbooks on selling and advertising by mail. Detailed instructions covering the lBull, op. cit., Ch. I, p. 7. 2Ibid., p. 9. 3Ibid., p. 11. 38 various aspects of mail order trade are found in several books of the mail order literature.1 More recently, attention has been given to the in-home retailing practices. The advent of technological innovations has pointed out manners in which such advance- ments could be used in marketing. The literature has been enriched by various articles on the subject in the last ten or five years. Mail order and in-home retailing experts try to transmit their past experience by providing thorough and complete manuals of instructions to the beginner in mail order business ventures.2 General In-Home Shoppipg_Literature Along with the literature which describes the his- tory of mail order selling as a part of the United States retailing system, and also with the literature concerned in providing rules and principles about how to operate and manage a mail order concern, studies and publications can be found about major topics closely related to the problems of in-home retailing. It is necessary to underline once 1Homer J. Buckley, G. D. Crain, Jr., and Maxwell Droke, Mail Order and Trade Paper Advertising (New York: International Textbooks Co., Bluebooks, 1937); and Frank Egner and L. Rohe Walter, Direct Mail Advertising and Selling (New York: Harper and Bros., 1940). 2Robert F. Stone, Successful Direct Mail Advertising and Selling (New York: Prentice-Hall, 1960); and James Everett Howard, How to Use Mail Order for Profit (New York: Grosset and Dunlap, 1963). 39 more that in-home retailing, as studied in this research, excludes all door-to-door sales, considering solely sales by mail, by telephone, or any other type of sales in which no direct contact existed between the buyer and the product, or a sales representative of the retailing firm. One of the major preoccupations of authors who have studied in-home shopping practices has been that of explor- ing different ways of selling by mail and by telephone, as well as the attempt to predict possible future changes in the technique of selling goods at home. This concern reflects the popularity that the in-home sales still enjoy among retailing practitioners and scholars. In-home shopping has been considered the fastest growing form of retailing by some authors.1 In other instances, in-home shopping has been regarded as a real retail revolution, something which would be enacted with the aid of sophisticated technology in the next ten to twenty years. The motive behind the retail revolution would be the effort to give more convenience to the customers.2 Many colorful ways of accomplishing this goal have been described, such as the use of house computer consoles to order products to a retailer who would be operating with lMaxwell Sroge, "The Distribution Revolution . . . Anatomy of Direct Marketing,” The Reporter of Direct Mail Advertising, July, 1970, pp. 18-20. 2Lester Wundermann, "Mail Order, The Coming Revolu- tion in Marketing, The Reporter of Direct Mail Advertising, Editor's Choice, Summer, 1968, pp. 32-34. 40 only one distribution center. Some reasons for this "inevitable" revolution are mentioned.1 Many housewives of the 1970's will have grown up with color television, push-button telephones, and computerized banking services. Indeed, these and other products of a highly sophisticated technology are already commonplace. The young matron of the 1970's probably will have a college education and have greater variety of interests and activities than her mother did. And she will enjoy greater economic security and comfort than did the woman of her mother's generation. All these things add up to the certainty that she will demand more and different services, and new marketing institutions will have to be developed to meet these demands.2 One new way of purchasing products has been called "tele-purchasing." It is defined as: ". . . Telepurchasing, or long distance buying in which the goods are again brought to the customer instead of the customer to the goods."3 The two major reasons listed which justify the arousal of in-home shopping as a more frequent activity than the current shopping patterns are: offering added convenience to the customer, so that he will use his leisure time in other activities of greater interest for him,“ and the fact that, by eliminating middlemen through this type of 1Alton F. Doody and William R. Davidson, "Next Revolution in Retailing," Harvard Business Review, May-June, 1967, pp. 4ff. 2Ibid., p. 16. 3"Tale-Purchasing, Major Trend in Retailing?" Forbes, October 15, 1967, pp. 56-64ff. l’Pat Terry, "More Customer 'Services' for Jordan March in 70's," Women's Wear Daily, March 26, 1970, p. 29. 41 distribution, the firm's profits would be larger because lower and more competitive prices could be afforded.1 Empirical evidence points out that the concern with segmenting markets existed for a long time within mail order merchants and scholars. Historically, one of the most pop- ular tools for attempting such strategy was the use of the so-called "mailing lists." A mailing list is a tentative measure to isolate a mail order merchant's core market. Mailing lists were compiled even at the earlier stages of the mail order business.2 More recently, authors have directed their attention to providing somewhat SOphisticated techniques to segment the mail order market. Different systems have been described in the literature which attempt to select lists according to variables such as sex, income level, type of occupation, and many others, indicating that emphasis is 3 given to possible submarket segments. Following the same 1Stanley Fenvessy, "The Dynamic Wholesale Mail Order Business: Evolution or Revolution?" The Reporter of Direct Mail Advertising, October, 1960, pp. 51-53; and "More Firms Selling Directly to the Consumer," Home Furnishing Daily, May 1, 1970, p. 7. 2Frank W. Sparth, "Classifying Direct Mail Lists," The Journal of Retailing, II (October, 1928), 13-17. 3Randall P. McIntyre, "Mail Marketing in Three Dimensions," The Reporter of Direct Mail Advertising, OCtOber,..l965, pp. 72-77. 42 rationale, retailers are invited to be "selective" as to the lists they use.1 List segmentation is at times approached in a more systematic manner. A good example is the ramification of several types of lists: . . . in direct mail there are three main sources of names, stratified lists, mail order lists, and compiled lists. In case of stratified lists the entire country is divided into micro-units, and householdslocated in each unit are characterized by census data. Mail order lists consist of people who have bought something by mail; and compiled data are those which are prepared on the basis of one or two characteristics--gen- erally from professional directories--e.g., lists of doctors, dentists, lawyers, school teachers, and executives.2 Ways of selecting customers according to their frequency of purchase,3 or according to break-even indexes per groups of customers“ are only two of the many examples of the preoccupation with segmenting in-home markets. The variables which are most commonly suggested as segmentation tools are primarily socioeconomic groups or combinations of them used as "indexes." 1Roy M. Green, "Think Creatively About the Lists You Use and The Markets You Sell," The Reporter of Direct Mail Advertising, June, 1967, pp. 16-17. 2C. L Jain and A. Migliaro, "The Fundamentals of List Segmentation," The Reporter of Direct Mail Advertising, June, 1967, p. 29. 3Jerome B. Osherow, "How to Clean a Mailing List," The Reporter of Direct Mail Advertising, October, 1960, pp. 39-40. I’Milton 8. Stevens, "How Market Research Can Improve List Results," The Reporter of Direct Mail Advertising, Editor's Choice, Summer, 1964, pp. 27-29. 43 Ed Burnett, who is an active writer on mail order and direct mail suggests that the economic variables are the most relevant variables for segmenting the in-home shopping markets. ". . . Direct mail insofar as merchandising of goods and services to consumers is concerned, can be described as a search for discretionary money."1 Attitudinal and psychological variables are not covered as extensively in the literature as segmentation tools for the in-home market. Seldom mention has been found of attempts to segment the mail order and telephone market by a combination of socioeconomic and sociopsychological variables. Also, it is very infrequent to find in the literature clear differentiation between markets for differ- ent types of in-home retailers. The list segmentation and the list selection pro- cedures described, therefore, don't show great concern with more sophisticated systematics than methods of stratifying samples by social or economic variables within geographic areas. Two other major subjects can be found discussed in the literature. A series of methods which teach the practitioner how to apply "mail order mathematics.2 1Ed Burnett, "How to Select Lists That Reach People With Ability to Buy," The Reporter of Direct Mail Advertis- ing, April, 1969, p. 34. 2Ray Snyder, "The Mathematics of Mail Order," The Reporter of Direct Mail Advertising, February, 1968, pp. 19-22. 44 Most of these are detailed applications of calculus like the "break-even analysis," or the develOpment of linear equations1 with the objective of maximizing profits or minimizing costs. Such articles are very similar, in their approach, to the initial "how to" literature in the early mail order retailing. Another type of information that can be found in the literature is the description of successful cases of mail order concerns. Applications of these retailing techniques to several different types of products are discussed in detail.2 References of this kind have merely historical value and do not report information relevant to marketing problems related to the in-home shopping activity. Social Class and Consumption Behavior Social class or socioeconomic status has been fre- quently accepted as a factor influencing the purchasing behavior for a large number of consumer's goods. Several researchers have attempted to establish a relationship between some social class classification and the consumption 1C. L. Jain, "The Cost and Value of Customers--How to Figure It for Yourself," The Reporter of Direct Mail Advertising, December, 1968, pp. 44-45. 2"The Diners Club Story," The Reporter of Direct Mail Advertising, Editor's Choice, Summer, 1967, pp. 18—22. 45 behavior.1 It cannot be denied that there are perceivable differences among strata of our pOpulation. It is also clear that such differences may be roughly evaluated such as to fit some scale measurement of values. A hierarchy of values and roles of individuals within society can be established leading to a social dif- ferentiation and classification; or in other words, to social classes. Engel, Kollat and Blackwell have defined social classes to be: "a relatively permanent and homoge- nous division in society into which individuals or families can be categorized where being compared with other individ- uals or families in the society."2 According to Howard, social stratification is attained by the roles that individuals play within society. He defines social stratification as: ". . . a structure of regularized inequality in which men are ranked higher and lower according to the value accorded their various social roles and activities."3 While individuals, because of their occupational activities and roles pertain to different social classes, 1John A. Howard, Marketing Theory (Boston: Allyn & Bacon, Inc., 1965), pp. l66ff; and Bernard Barber, Social Stratification (New York: Harcourt, Brace & World, Inc., 1957). p. 2. 2James F. Engel, David T. Kollat and Roger D. Black- well, Consumer Behavior (New York: Holt, Rinehart & Winston, 1968), p. 264. ”John A. Howard, op. cit., pp. 166-167. 46 other variables as education, family background and cultural diversions or even religious and economic behavior interact to form common patterns which perpetuate and strengthen societal hierarchy: Social classes are the vehicles through which the totality of culture is made specific for a family and ultimately an individual. Social classes define the expectations of society for groups of people and for families within the groups. The family then transmits these cul- tural expectations to the individual. Social classes create different patternings of behavior specific to groups of peOple but drawn from a common and pervasive group of elements in the core culture. The variations among classes in cultural manifestations may be subtle or they may be obvious, and the marketer seeks to understand what these differences are.1 There are many studies in the literature which attempt to measure the effects of social classes on the purchasing behavior of individuals or groups of individuals. The researcher will try to examine in part what has been done in this field. The first problem posed to the scholars has been to define and somehow measure social classes in order to clas- sify individuals according to them. Several independent variables have been indicated as determining social classes. Joseph A. Kahl indicated that social classes have six major parameters: (1) social interaction, (2) occupation, 'Engel, Kollat and Blackwell, op. cit., p. 267. 47 (3) personal prestige, (4) possessions, (5) class consciousness, and (6) value orientations.1 Other dimensions of social classes have been added. Among the most frequently mentioned by the authors are: (1) authority, (2) power, (3) ownership of property, (4) consumption patterns, (5) education, (6) morality, (7) kinship and ancestry, (8) associational ties, and (9) ethnic status.2 The amount of interest devoted to this area of marketing has produced a large number of studies about social class. Several of the methods which attempt to measure an individual's social status are listed below. The Reputational Method The reputational method consists of ranking indi- viduals by other individuals who are familiar with these, on a continuum of social scales.3 The smaller the community and the more knowledge peOple have about each other's char- acteristics and habits, the easier it becomes to obtain such classification. lJoseph 1L. Kahl, The American Class Structure (New York: Holt, Rinehart and Winston, Inc., 1957), pp. 8-10. 2Edward A. Shils, "Class" in the Encyclopedia Britannica, Vol. V, 1960, pp. 766-768. 3Engel, Kollat and Blackwell, op. cit., p. 264. 48 The reputational method can be divided into the subjective and the objective techniques. With the subjec- tive technique reSpondents are asked to classify themselves into a social class. The results of this method depend on the relative class awareness of each individual. Research on this tOpic was made by John L. Haer in Tallahassee, Florida.1 Results showed that, in three different samples between 35.5 and 53 percent of the respondents either could not place themselves in a social class group or did so in an incorrect manner. This suggests, according to the author, that the "latent" class orientation simply may not exist for large portions of the pOpulation.2 The objective technique involves observing individ- ual behavior to determine their respective social class. Several indexes were developed for classifying individuals according to their socioeconomic status. W. Lloyd Warner was responsible for the development of six reputational techniques to measure social class. These techniques are: rating by matched agreements, rating by symbolic placement, rating by status reputation, rating by comparison, rating by simple assignment to a class, and rating by institutional membership. In these techniques the analyst devises rating mechanisms which will translate the 1John L. Haer, “An Empirical Study of Social Class Awareness," Social Forces, XXXVI (December, 1957), 117-121. 2Ibid., p. 120. 49 criteria and judgments of the informants into explicit results which will correSpond with the class realities of the community.1 Warner's approach has not been widely adopted by marketing researchers, primarily because of its complexity.2 Hollingshead develOped an objective method for classification of socioeconomic status. In his study he used weighted combinations of three variables: residence, occupation and income. His reasoning for using three vari- ables was that the family's mode of living is "mirrored in its home," and that occupational level reflects the skills and social power of an individual, as well as the formal education reflects an individual's tastes.3 With this method, Hollingshead established five major social classes: 1. "Old families," top business management and profes- sional occupations, high income, highly educated, expensive homes, the social elite. 2. Business managers, lesser professionals, often college graduates, socially sensitive, "on the way up. 1W. Lloyd Warner, Marcia Meeker and Kenneth Eells, Social Class in America (New York: Harper Torch Books, Inc., 1960): PP. 37-38. 2Engel, Kollat and Blackwell, op. cit., p. 276. 3A. B. Hollingshead and F. C. Redlich, Social Class and Mental Illness (New York: John Wiley and Sons, Inc., 1958). 50 Employees in various salaried administrative pursuits, small business owners, average incomes, high school graduates, modest homes in "good" areas. Semiskilled and skilled manual employees, below average incomes, many are homeowners but live in multiple units, many had some high school but did not graduate, often members of minority ethnic groups. Unskilled and semiskilled, low incomes, live in old tenement areas, most did not finish grade school, "live today, let tomorrow take care of itself."1 The Sociometric Method The sociometric method measures social classes by inquiring from an individual the references about his close associates, and successively through observation of the individual's activities as well as those of his acquaint- ances.2 suming. This method is complex, expensive and time con- As an example, a research made by Hollingshead using this technique required that he, his wife and a fellow researcher live for nineteen months in the city where the study was made.3 'Ibid., p. 211. 2Engel, Kollat and Blackwell, op. cit., p. 216. 3A. B. Hollingshead, Blmtown's Youth (New York: John Wiley and Sons, Inc., 1949). 51 The Occupational Method The occupation of an individual has been widely studied as a variable directly related with his social status. William L. Hurst began segmenting occupations into classes as early as 1897.1 However, occupation as a measure of social class was accepted completely only in 1917 when Alba M. Edwards introduced what he called "social economic grouping."2 Edwards' scale is still widely used. However, as it has shown to correlate only slightly with average levels of income and education, it should be adOpted only where a very rough indicator of social class is needed.3 Two other scales of occupation have been developed with a more specific and clear approach than that used by Edwards: the Duncan scale which ranks 425 occupations,“ and the scale prepared by the National Research Center with 90 occupations.5 lU.S. Bureau of the Census, Methodology and Scores of Socioeconomic Status, 1960, p. 1. 2Ibid., p. 1. 3Ibid., p. 2. “Albert J. Reiss, Jr., Otis Dudley Duncan, Paul K. Halt, and Cecil C. North, Occupations and Social Status (New York: The Free Press, 1961). 5Robert W. Hodge, Paul M. Siegel, and Peter H. Rossi, "Occupational Prestige in the United States," in Class, Status and Power, ed. by Reinhard Bendix and Seymour Martin Lipset (2nd ed.; New York: The Free Press, 1966), pp. 322- 334. 52 The Multiple Index Method The Multiple Index method consists of constructing social class measures taking into account more than one determining variable. The objective of this method is to avoid errors or distortions that could occur where only one variable is utilized. Lloyd Warner's Index of Status Characteristics is one of the most pOpular multiple indexes used to determine social class. Warner uses four variables to compose his index: occupation, source of income, house type and dwell- ing area. Each one of these variables is weighted and the weighted total indicates the social class affiliation of the individual on the scale.1 The weights for the factors were obtained by regression analysis and indicate the relative importance of each factor. Warner states that three of the factors may be used to measure social class, however he recommends that, whenever possible, all four factors are used. Another very well known multiple index for rating individuals according to their social classes, is that develOped by the U.S. Bureau of the Census.2 The variables used to compare the index are: occupation of the head of the household, formal education of the head of the house- hold, and total family income. The weight of each variable lWarner, Meeker and Eells, op. cit., pp. 37-38. 2U.S. Bureau of the Census, 1960, op. cit. 53 in this index is the same. Scores are given to 494 job classifications, to all possible combinations of years of school completed and to categories of family income. The scores for each of the three variables are added and divided by three. The income score is the individual's socioeconomic status score. A comparison was made by various indexes of socio- economic status in a study develOped by Joseph Kahl and James Davis.1 The purpose of the study was to verify if there was a relationship among various indexes of socio- economic status. If this occurred, then it would be an indication of the validity of a measure like "social class." The authors underline that: The research man . . . can tentatively con- clude that "socioeconomic status" is an accurate though clumsy term: there is a composite of social and economic attributes that tend to cluster together, and we can measure the com- posite fairly well. For many purposes it is practical to treat this composite as one dimen- sion--the general factor. The best single index of it is an occupational scale. (Warner's has a higher loading, but the Census Bureau gives details that make coding more reliable.) Some improvements in measurement can be had by com- bining occupation and one of the variables from the second common factor, such as Census tract, with a heavier weight given to the former, to add more variable to a composite index."2 1Joseph A. Kahl and James A. Davis, "A Comparison of Indexes of Socio-Economic Status." Social Forces, May,l960. 2Ibid., p. 321. 54 Kahl and Davis used factor analysis to list 19 stratification indexes for 219 men. Results showed that the indexes were highly correlated because all measured, roughly, the same underlying dimensions. A conclusion can be drawn from the study: The battery of indexes showed two common factors. The first was composed of the various measures of occupation, plus certain variables closely related to occupation, such as education self-identification, and the interviewer's rating of the subject. The second factor was composed of ecological measures plus those of the status of the parents of the subject and his wife. The two common factors accounted for most of the mutual variances of the original indexes. The little that remained was studied in a cluster analysis. It revealed certain sub-groupings of variables that were highly related to each other after their mutual relationships to the common factors were controlled. These turned out to be mainly clusters of indexes that were replicating measures of the same variable, a further indica- tion of the fact that the two common factors "explained" most of the important relationships among the 19 indexes.1 The Application of Social Classes to Marketing Researchers have studied the relationships which exist between the fact that an individual being is clas- sified as belonging to a social class and his purchase patterns or habits. Pierre Martineau is responsible for exercising a strong influence in focusing the marketer's attention on social classes as an important variable to explain con- sumers' behavior. 'Ibid., p. 325. 55 In his article "Social Classes and Spending Behavior Martineau reports findings of a study undertaken by W. Lloyd Warner and himself for the Chicago Tribune.1 The study objectives were: to determine if such a social class system did exist in metropolitan Chicago, if the dimensions and the relationships were at all similar to the smaller cities which were studied before the far- reaching social changes of the past fifteen years. The studies were undertaken to see if there were any class significances in the individual family's spending-saving patterns, retail store loyalties, and his expression of taste in typical areas such as automobiles, apparel, furniture and house types.2 Martineau and Warner's study was made with a sample of about 4,000 households in the metrOpolitan area of Chicago. The researchers used an Index of Status Character- istics which enables them to divide the sample into five distinct social classes similar to Warner's classification. The findings were significant as to the variables studied. They revealed a close relationship between choice of store, patterns of spending and class membership. People are very realistic in the way they match their values and expectations with the status of the store. The woman shopper has a considerable range of ideas about department stores; but these generally become organized on a scale ranking from very high social status to the lowest status and prestige. The social status of a department store becomes the primary basis for its definitions by the shopper. This is also lPierre Martineau, "Social Classes and Spending Behavior," Journal of Marketing, October, 1958, pp. 121-130. 2Ibid., p. 122. 56 true of men's and women‘s apparel stores, and furniture stores, on the basis of customer profits. The shopper is not going to take a chance feeling out of place by going to a store where she might not fit.1 The communication ability also showed a relation with social classes. Lower status individuals prefer simple, unsophis- ticated messages to the subtle humor of advertising which is carried by such magazines as the New Yorker and Esquire. The type of communication they are used to employ differs substantially from the symbols which are commonplace among high class individuals. Here again, style of advertising helps the individual to make class identification. Most of the really big local television success stories in Chicago have been achieved by personalities who radiate to the mass that this is where they belong. The self-made businessmen who do the announcing for their own shows communicate wonderfully well with the mass audience. While many listeners switch off their lengthy and personal com- mercials, these same mannerisms tell the lower status individual that there is some- one just like himself, who understands him.2 Social class status was also found to influence the spending and saving patterns of individuals. The higher status individuals showed more savings aspirations than the lower status ones. Also, the manner by which savings would be employed was more Specifically defined by higher status people. 1Ibid., p. 126. 21bid., p. 128. 57 Intangible forms of investment like stock and insurance are very clearly distinguished as upper status investments. Nearly four times as many upper middles select insurance as would be expected by chance, whereas only one-fifth of the lower-lowers select it as would be ex- pected by chance. By contrast, lower status people have far greater preference for tangible investments, specifically ownership of real estate, a farm or a business.1 Warner and Martineau's research showed that con- sumption habits can be a status symbol which indicates or defines class affiliation. They conclude that class membership is a more significant determinant of economic behavior than income. The research proved the relevance of the social class concept for marketing problems relating to the purchasing behavior of individuals. Coleman makes a second contribution in the area of social class and consumption behavior. He conceptualizes as overprivileged those individuals whose income is higher than average for their class; he calls underprivileged those individuals whose income is below the average income for their class. Coleman states that the purchasing behavior of underprivileged and overprivileged individuals did not match that of the individuals whose income was typical of their social class.2 'Ibid., p. 128. 2Richard P. Coleman, "The Significance of Social Stratification in Selling," in Proceedings of the 43rd National Conference of the American Marketing Association, ed. by Martin L. Bell, December, 1960, pp. 171-184. 58 Social class variations are variations in life-style. Although social class groups are not sharply distinguished by their behavior in most studies, they do show behavior that can be viewed as ranging along a continuum or as dif- ferent patternings using common elements drawn from the same global culture. To many marketers those differences are important and useful. One of the most important advantages of socioeconomic status classification is the fact that it enables segmentation of the market. In the research, social class will be used as an independent variable in differentiating the total retail market into in-home shoppers and not in-home shoppers. Also it will be used to segment the various submarkets within the in-home shopping market. Family Life-Cycle and Consumppion Behavior It has been observed that families behave in dif- ferent ways according to their stages in life. These differences in behavior have been shown to be of great interest also for economists and marketers. The study of family life-cycle began as a consequence of the urban migration. Rural sociologists, interested in verifying the differences between rural and urban cultures, were the first to study the phenomenon.1 1Pitirin A. Sorokin, Carl C. Zimmerman and Charles J. Galpin, A Systematic Source Book in Rural Sociology (Minneap- olis: University of Minnesota Press, 1931), Vol. II, pp. 3- 32; and Charles P. Loomis, "The Study of the Life Cycle of Families," Rural Sociology, June, 1936, pp. 180-199. 59 A series of articles appeared during the early 1930's,1 which analyzed family life-cycle of rural commu- nities. These studies used a four stage family life cycle based either on the criterion of expanding and contracting family size or on the age of children. It was only in 1947 that the first analysis of U.S. Census material was published.2 In this study family life- cycle was divided into seven stages based upon a statistical analysis of median values. Also during this period appeared the first studies which attempted to indicate the importance of family life-cycle for some economic characteristics.3 In 1948 a report by Duvall and Hill at the National Conference on Family Life held in Washington, D.C.,“ analyzed family life-cycle with specific reference to 1C. E. Lively, "The Growth Cycle of the Farm Family" (mimeographed bulletin No. 51; Wooster: Ohio Agricultural Experiment Station, 1932); C. P. Loomis, "The Growth of the Farm Family in Relation to Its Activities" (Raleigh: North Carolina State College, Agricultural Experiment Station, 1934); and E. L. Kirkpatrick, Mary Cowles and Roselyn Tough, "The Life Cycle of the Farm Family" (Research Bulletin 121; Madison: University of Wisconsin, Agricultural Experiment Station, 1934). 2Paul C. Glick, "The Family Cycle," American Sociological Review, XIV (April, 1947), 164-174. 3Howard F. Bigelow, "Financing the Marriage," in Family, Marriage and Parenthood, ed. by Howard Becker and Reuben Hill (Boston: D. C. Heath, 1948), pp. 393—418. “Evelyn M. Duvall and Reuben Hill, Co-Chairmen, Report of the Committee on the Dynamics of Family Interac- tion (Washington, D.C.: National Conference on Family Life Cycle, February, 1948), mimeographed. 6O interactive processes. In this report, a seven stage cycle was described analyzing the processes of interaction which appeared in the family at various periods. In 1957 Click published American Families,1 where he made a detailed study of the United States population comparing data from 1890 and the 1950 censuses. The study is primarily a description of the social changes which occurred in the family composition within the period considered. Also in a 1957 study on the economic behavior of families was published by Lansing and Kish.2 The conclusion of their research was that the stages of the family life- cycle were a very significant explanation of the age classes of the head of the household. Lansing and Kish tested family life-cycle relationship with variables such as owner- ship, indebtedness, employment of wife, income level, pur- chase of an automobile, and purchase of a television set. They found that those variables were related to the family life-cycle. More recent additions to the study of family life- cycle are two books edited by Foote and others.3 In Housing 1Paul C. Glick, American Families (New York: John Wiley and Sons, 1957). 2John B. Lansing and Leslie Kish, "Family Life Cycle as an Independent Variable," American Sociological Review, XXII (October, 1957), 512-519. 3Nelson N. Foote et al., eds., Housing Choices and Housing Constraints (New York: McGraw-Hill, 1960); and Nelson N Foote, ed., Consumer Behavior: Models of Household Decision Making (New York: New York University Press, 1961). 61 Choices and Housing Constraints, we find a study of the relationship between type of housing and stages in the family life cycle, while Consumer Behavior: Models of Household Decision Making, applies the developmental theory to various issues of the decision-making process by the family. Those two books, as well as some of the preceding studies do not develOp theories about family life—cycle, they merely use it as an independent variable in order to explain other phenomena, setting a trend in the current literature. Techniques oprlassification of Family Life-Cycle Besides the generalistic attempts to classify stages in the family develOpment and which limited themselves pri- marily to a description of stages of contraction and stages of expansion, specific efforts were made to obtain narrower and more specific classifications. The study by Glick, in 1947,1 was one of the pioneering tentatives of classifica- tion. Glick developed seven steps examining variables such as: family size, dissolution of the family, children living at home, adult relatives in the home, residential shifts, home ownership, family income, and employment of the husband and wife. lGlick, American Families, op. cit. 62 Evelyn Duvall is responsible for another study designed to develOp a classification for family life—cycle.l The criteria that Duvall used to determine the definitions of life cycle categories are: (l) plurality patterns, (2) age of oldest child, (3) school placement of oldest child and (4) functions and statuses of families before children came and after they leave. With these criteria, the outlines are eight stage family life cycles defined as follows: Stage I. --Beginning families (married couple without children). Stage II. --Childbearing families (oldest child, birth to 30 months). Stage III. --Families with preschool children (oldest child 2-1/2 to 6 years). Stage IV. --Families with school children (oldest child 6 to 12 years). Stage V. --Families with teenagers (oldest child 13 to 20 years). Stage VI. --Families at launching centers (first child gone to last child's leaving home). Stage VII. --Families in the middle years (empty nest to retirement). Stage VIII. --Aging families (retirement to detah of one or both spouses).2 Duvall, however, underlines the problem of existence of overlapping stages in her classification. In defining stages in terms of oldest child, the presence of other children in the family is not explicitly recognized. A clear-cut sequence of stages of family life cycle such as that outlined above seems to occur only 1Evelyn M. Duvall, Family Development (New York: J. B. Lippincott Company, 1947). 2Ibid., p. 8. 63 in the one-child family. In families of more than one child, there are several years of overlap at various stages.1 A little further, Duvall attempts a solution of the problem: There is no simple solution to the conceptual problem of overlap of stages of family life cycle in families of more than one child. Since our thesis is that families grow and develOp as their children do, our answer to the question of overlapping of stages is that a family grows through a given stage with its oldest child, and in a sense "repeats" as subsequent children come along. In order to correct the problem of overlapping stages, Roy H. Rodgers suggested a third type of clas- sification: l. Childless couple. 2. All children less than 36 months. 3. Pre-school family with (a) oldest 3 to 6 and youngest under 3, (b) all children 3 to 6 years. 4. School-age family with (a) infants, (b) pre-schoolers, (c) all children 6 to 13 years. 5. Teenage family with (a) infants, (b) pre- schoolers, (c) school ages, (d) all chil- dren 13 to 20 years. 6. Young adult family with (a) infants, (b) pre-schoolers, (c) school ages, (d) all children 13 to 20. 7. Launching family with (a) infants, (b) pre- schoolers, (c) school ages, (d) teenagers, (e) youngest child over 20. 8. Where all children have been launched until retirement. 9. Retirement until death of one spouse. 10. Death of first spouse to death of survivor.2 1Ibid., p. 9. 2Roy H. Rodgers, "Improvements in the Construction and Analysis of Family Life Cycle Categories" (unpublished Ph.D. dissertation, Western Michigan University, Kalamazoo, 1962). 64 Rodger's approach is more objective and definite. It seems to have avoided as much as possible the problems of overlapping stages. Consumption and Family Life-Cycle The concept of family life-cycle was used as an independent determining variable in several studies which attempted to predict consumer behavior patterns through the analysis of his stage in the family life cycle. Of such nature are studies by Kirk and Reid.1 In 1958, a study by Braid related family life cycle with such variables as interdependence of the family status, personal income and size of the consuming unit.2 Lansing and Morgon published a very complete study which attempts to relate consumer behavior with family life cycle.3 They found very relevant data which explain some of the consumption phenomena related to the stage in the family life-cycle. Wilton Thomas Anderson, Jr. in his Ph.D. disserta- tion summarizes all attempts that have been made to link family life cycle and consumption: 1Hazel Kirk, A Theory of Consumption (Boston: Houghton Mifflin Company, 1923); Hazel Kirk, Economic Problems of the Family (New York: Harper and Row, Publish- ers, 1933); and Margaret Reid, Consumers and the Markets (New York: Appleton-Century-Crofts, 1938). 2Dorothy Braid, "Individual Incomes and the Struc- ture of Consumer Units," A.I.A. Papers and Proceedings, May, 1958. 3John B. Lansing and James N. Morgon, "Consumer Finances Over The Life-Cycle," in Consumer Behavior II: The Life-Cycle and Consumer Behavior, ed. by Lincoln H. Clark (New York: New York University Press, 1955). 65 The interpretation of attempts to characterize specific dimensions of the family consumption quotient erects an economic conceptual frame- work grounded in three basic concepts. 1. Standard of living and their dependence upon custom, habit, emulation, tastes, and the means for financing consumption: income and occupation. 2. Socioeconomic status and its numerous indices ofirank: occupation, source of income, education, housing, and material goods and services, that is, the products of consumption. 3. Consumer behavior and the motivating forces behind consumption: wants and needs, utility and value.1 Personality Variables The focus of this thesis is to determine whether the in-home shOpping market can be segmented considering person- ality variables as well as socioeconomic status and family life-cycle. As has been stated earlier in this chapter, the great majority of list segmentation studies have been con- cerned with socioeconomic variables. No studies have been found which use personality variables to differentiate in- home shOppers, or to distinguish among different types of in-home shOppers. The merits of personality measures as a tool for market segmentation have been eXpressed by several behav- iorists. Engel, Kollat and Blackwell explain that lWilton Thomas Anderson, Jr., "An Analysis of the Correlates of Convenience Orientated Consumer Behavior: With Special Emphasis on Selected Convenience Foods and Durable Goods" (unpublished Ph.D. dissertation, Michigan State University, 1969), p. 86. 66 personality may be more useful for segmentation when used along with other variables: Personality is useful in a different way if the market is first segmented by some other subjective mean. The subgroups isolated are then analyzed to detect any differences in psychological attributes. In other words, personality is no longer the independent variable used to segment the market; rather, it is used along with other measures to determine differences between members of segments isolated by other means.1 Another recent study developed by Myers and Alpert is concerned with attitudes as determinants of buying behavior.2 The authors define as "determinant attitudes": "the attitudes toward features which are most closely related to preference or to actual purchase decisions."3 They agree that marketing strategy must be built around the determinant attitudes. This study proposes methods of evaluating such attitudes and contributes to enlarge and specify the differences between personality character- istics and attitudes as influencing variables of the buying process. 1James F. Engel, David T. Kollat and Roger D. Black- well, "Personality Measures and Market Segmentation-- Evidence Favors Interaction Views," Business Horizons, June, 1969, pp. 61-69. 2James H. Myers and Mark I. Alpert, "Detrimental Buying Attitudes: Meaning and Measurement," Journal of Marketing, XXXII (October, 1968), 13-20. 3Ibid., p. 13. 67 Another study of the relationship between consumers' attitudes and their buying behavior was developed by Udell.1 He investigated the influence of attitudes on the activity of buying from retailers who had given trading stamps. Udell found that there was a relationship between a favor- able attitude towards trading stamps and the fact that they were saved by those interviewed. Frederick E. May summarizes in an article published in 1965 a great part of the results achieved through empir- ical research on buying behavior of household customers.2 Although May's list is not complete, it gives a very sig- nificant overview of the importance in marketing of inves- tigating variables related to the consumer personality characteristics as well as social influences which might determine or define his buying patterns. The choice of the personality and attitudinal variables used in this segmentation study were a result of personal meditation about the specific problem. Consider- ation was given to the "total product" and the possible reasons for its purchase, to the type of distribution along with other variables which might interact to provoke the purchase. It is reasonable to eXpect that other variables 1John G. Udell, "Can Attitude Measurement Predict Consumer Behavior?" Journal of Marketing, XXIX (October, 1965), 46-50. 2Frederick B. May, "Buying Behavior: Home Research Findings," The Journal of Business, XXXVIII (October, 1965), 68 not considered in this research might be as significant or even more significant than the ones included in the study. The investigation of buyer behavior is far from being complete. Only further tentatives will prove the extent and validity or personality variables as a tool for seg- mentation of the in-home shopping market. Segmentation of the In-Home ShOpping Mgrket by Socioeconomic and SoEIOpsychological Variables As it was indicated in Chapter I and previous segments of Chapter II, there is clear evidence of growth of the in-home shopping market. The importance of this activity has been shown by the existence of extensive literature on the subject. As a market grows larger and more important, greater attention is given to its components and there is generally a proliferation of competitive efforts to gain the prefer- ence of consumers. As it was mentioned in Chapter I, such marketing efforts have been translated into various strat- egies, one of the most important of which is the strategy of market segmentation. Several criteria have been used in the past by marketing practitioners and scholars in order to segment markets. The researcher has shown how, in the literature, studies of segmentation have been reported which have used separate variables such as: social class, stage in the family cycle and personality characteristics. 69 No trace has been found, however, of a research study which analyzes in—home shopping market segments by investigating these three characteristics of shoppers combined: their social class, along with their stage in the family life-cycle and their personality characteristics. Since the relevance and importance of each one of these variables for marketing is accepted by the past literature, it seemed natural that a combined study of these would add valuable information to marketing knowledge and practice. The conceptual parameters within which the present research has been planned are eXpressed by the inter- relationships which are assumed to exist between in-home shopping activities and trends; and the socioeconomic and sociopsychological characteristics of in-home shOppers. This study attempts to synthetize jprevious efforts of segmentation in order to recognize with more detail the existence of submarkets within the total in-home shopping market. The methodology utilized in the research is explained in detail in Chapter III. CHAPTER III RESEARCH DESIGN Chapter III is concerned with the research design of the study. In the first portion of the chapter the dependent and independent variables are presented. Follow- ing is a discussion of sample design and procedures used to collect and analyze the data. A complete c0py of the ques- tionnaire used in the research is attached as Appendix C. Research Design Framework Identification of Variables The methodology used in the study reflects the objectives of determining socioeconomic and sociopsycholog- ical characteristics of in—home shOppers.l The respondents' answers as to the frequency of their in-home purchases served as a criterion to differentiate between those indi- viduals who do and do not shop at home. Their answer as to what type of in-home retailing outlets they used to purchase lIn-home shOppers are defined as individuals who purchase products either by mail, or by telephone, or both. PeOple who purchase products from door-to-door salesmen are not considered to be in-home shOppers in the present research. 70 71 their products served as a classification instrument to determine the different segments of the sampled in-home shOpper's market. The research design was limited to the delineation of different patterns of behavior of in-home shoppers and individuals who do not shOp at home. Socioeconomic and sociopsychological characteristics of respondents were identified as independent variables. Patterns of in-home shopping and frequency of in-home purchases were identified as dependent variables. Independent Variables Two sets of independent variables were examined in the study. The first set consisted of socioeconomic and demographic variables. The purposes of using these vari- ables were to measure the respondents' social class,1 and their stage in the family life-cycle.2 Specifically, the variables concerning social class and life cycle character- istics of the respondents in the research were the following: (1) marital status of the housewife, (2) employment status of the housewife, (3) age of the head of the household, (4) number and ages of children living at home, (5) occupation lU.S. Bureau of the Census, U.S. Census of Popula- tion, 1960, Subjects Report Socioeconomic Status, Final Report PL(2)-5c (Washington, D.C.: Government Printing Office, 1967). 2William H. Cunningham and W. J. E. Crissy, "Market Segmentation by Motivation and Attitude," Journal of Market- ing Research, IX (February, 1972), 100-102. 72 of the head of the household, (6) total family income and (7) level of education of the head of the household. The second set of independent variables consists of six sociopsychological scales. These scales attempt to measure the respondents' following personality and atti- tudinal characteristics: (1) trust in people, (2) cosmo- politanism, (3) attitude toward credit, (4) tendency to impulse buying, (5) adventuresomeness, and (6) conserva- tism. Each of the six scales are discussed below. Trust in people.--The trust in peOple scale consists of questions 29-33 in Section I of the Questionnaire (see Appendix C). The first three items were develOped by the 1 while Survey Research Center at the University of Michigan, the remaining two items came from a study by Rosenberg.2 The five items were scored in the Likert manner. The Likert scoring technique involves summing the numerical scores of each item in a particular scale. This sum then represents the respondent's score on the respective variable. Cosmopolitanism.--The cosmopolitanism scale is made up of the first six items in Section II of the Questionnaire. 1John P. Robinson and Phillip R. Shaver, Measures of Psychological Attitudes, Survey Research Center, Institute for Social Research, University of Michigan, Ann Arbor, MIchigan, 1961, pp. 529-532. 2M. Rosenberg, Occupations and Values (Glencoe, Illinois: The Free Press, 1957), pp. 25-35. 73 The scale was developed by Campbell in 1966 and was first reported as an unpublished working paper of the Department of Psychology, Northwestern University.1 This variable was also scored in a Likert manner. Attitude toward credit.--The researcher was unable to locate a test that did an adequate job of measuring individuals' attitude toward credit. As a result, items seven to twelve in Section II of the Questionnaire were designed specifically for the present study to measure this particular variable. Although only a limited amount of pre- testing was done, it is felt that the six item Likert test does give at least a general indication of the respondents' attitudes toward credit. The pretesting of the attitude towards credit scale consisted of first asking a panel of five doctoral students at Michigan State University to answer the questions and to examine them from the perspective of their own feelings toward credit. The second step in the pretesting process involved asking 25 randomly selected individuals from the greater Lansing community to complete the entire question- naire and then to comment on its readability and clarity. The same pretesting procedure was used for the tendency for impulse buying and adventuresomeness scales as well. 1Robinson et al., op. cit., p. 648. 74 Tendency for impulse buying.--Questions thirteen to eighteen in Section II of the Questionnaire were developed for the study in an attempt to elicit the respondents' attitudes toward impulse buying. Limited pretesting of the scale was done, in the same manner as for the attitude toward credit scale. It is felt that the five Likert items give an indication of the respondents' attitudes toward impulse buying. Adventuresomeness.--The adventuresomeness scale was developed for the research project; it consisted of questions nineteen to twenty-four in Section II of the Questionnaire. The six items were scored in a Likert manner. Conservatism.--The conservatism scale was developed by McClosky and was first reported in the American Political Science Review.1 The scale was designed to measure general conservative attitudes rather than just political attitudes. The test consists of the last ten items in Section II of the Questionnaire, and was scored in a Likert manner. Dependent Variables The first two objectives of the research, as stated in Chapter I of the study are: (1) to determine if there are significant differences among peOple who shOp at home, lHerbert McClosky, "Conservatism and Personality," American Political Science Review, 1958, No. 52, pp. 27-45. 75 and (2) to determine whether relatively distinct market segments can be identified within the in-home shopping market. The first eleven questions in Section I of the Questionnaire are designed to obtain information concerning the respondents' shopping patterns. Each subject is ini- tially classified as purchasing products at home either regularly, occasionally, very rarely or not at all.1 In order to determine if the subjects who shopped at home by mail or by telephone could be differentiated from those who did not, the respondents who indicated that they regularly or occasionally shopped at home were grouped to- gether as were the individuals who stated that very rarely or never purchased goods at home.2 The first respondent group is referred to as active in-home shOppers and the second group as inactive in-home shoppers. It was felt that, by contrasting the active and the inactive in-home shoppers, it would be possible to draw certain conclusions which would be generalized to the sample as a whole. It was also felt that it would be important to determine whether the subjects who indicated they were 1Question #7, Section I of the Questionnaire. 2Precedents for dichotomizing the respondents in this manner can be found in J. Hawey, "What Makes a Better Seller," in Motivation and Market Behavior, ed. by R. Ferber and H. G. Wales (Homewood, Ill: Richard D. Irwin, Inc., 1958), pp. 361-381; and W. T. Tucker and John J. Painter, "Personality and Product Use," Journal of Applied Psychology, No. 45, October, 1961, pp. 325-329. 76 regular in-home shOppers could be differentiated from those who stated they never shopped at home. Therefore, the regular in-home shoppers and those who never shopped at home were separated from the remainder of the sample and were analyzed to determine if significant differences existed between them. In order to define the existence of segments within the in-home shOpping market, the respondents were asked to indicate from what type of in-home shOpping outlet they purchased the goods.1 As no detailed classification of in-home sellers was found in the literature, for purposes of facilitating their answers as to where they purchased their products, in Section I of the Questionnaire, a tenta- tive enumeration of possible alternatives, was included. The criteria used to establish this enumeration was empir- ical observation of differences among in-home retail orga- nizations as to the type of products sold and the conditions of sale offered to the consumer. The six major categories of in-home retailers which appear in the Questionnaire are: (1) large catalogue and department stores, (2) "Club-of-the- Month” organizations, (3) novelty catalogue organizations, (4) specialty houses, (5) newspaper and magazine advertise- ments, and (6) credit card organizations. lSee questions 8 and 10, Section I of the Questionnaire. 77 Sample Selection The data were collected in June and July, 1970, in the cities of Lansing and East Lansing, Michigan. The cities were divided into the following five income tracts: (l) $6,000-$8,000, (2) $8,001-$10,000, (3) $10,001-$12,000, (4) $12,001-$14,000, (5) $14,001-$16,000. Two hundred families were randomly selected from each income tract to participate in the research. This was done in an attempt to insure that a cross section of the population in the urban test area would be included in the sample. Ed Burnett, in an article on potential buyers of mail order retailers states that: Discretionary purchasing power--the dollars most of us, as consumers, can spend as we please, after meeting truly basic needs, is a major fac- tor in the success of direct mail. And probably from the day some unsung genius recognized that letters, sent in quantity, could sell, the direct mail marketer has attempted to reach so-called 'peOple of means.‘1 This statement indicates that income is considered an important variable as far as influencing in-home pur- chases. The sample was then,'stratified, taking into account this particular factor. As Ferber, Blankertz and Hollander explain in their text on marketing research, a stratified sample is: "a lEd Burnett, "How to Select Lists that Reach People with Ability to Buy,” The Reporter of Direct Mail Advertis- ing, April, 1969, pp. 34—42. 78 1 The authors weighted aggregate of unrestricted samples." proceed in indicating what are the conditions which have to exist in order to use a stratified sample: 1. Certain relevant characteristics which influence strongly the subject under study must be known. 2. Division of the population by the relevant characteristics must be practicable. 3. The relative division of the population by these characteristics must be known with a fairly high degree of accuracy.2 Accordingly, as it was assumed by the researcher that income level of individuals was a relevant character- istic which influenced in-home shopping, the first require- ment was satisfied. On the same line, the sample was stratified by income level by Census Tracts, which indicated the practicability of the second requirement and also the existence of the third requirement. As the number of individuals selected from each income tract was not prOportional to the relative size of the stratum of the pOpulation, the sample used is defined as a disproportionate stratified sample.3 It was decided to use the disprOportionate sample because the various strata were not homogeneous with respect to the character- istic under study. lRobert Ferber, Donald F. Blankertz and Sidney Hollander, Jr., Marketing Research (New York: The Ronald Press Company, 1964), p 205. zIbido I pp. 205-2060 3Ibid., p. 206. 79 The stratified sample, when all the conditions for its use are present, has some advantages over the unre- stricted random sample. As Ferber, Blankertz and Hollander indicate: The sampling error of a stratified sample may be considerably below that of an unrestricted sample of the same size. This is because the stratified design makes use of the additional information regarding the breakdown of the population by the relevant characteristic(s) error. It is important to add that the questionnaire used in the research was anonymous, therefore, it is impossible to tell the number of questionnaires which were returned from each income tract. This fact does not alter the use of a stratified sample.2 The families were sent a set of materials which included a letter, a questionnaire and a pre-addressed stamped envelope. The letter indicated that the family had been selected to participate in a study sponsored by Mich- igan State University and that their cooperation would be greatly appreciated (see Appendix A). The families were urged to complete the questionnaire at their earliest con- venience. Two weeks after the first set of materials had been mailed, all the families in the sample were sent a post-card containing further requests to answer the ques- tionnaire and send it back to the researcher (see Appendix B). 'Ibid., p. 206. zIbid. 80 Six hundred questionnaires were returned, 81 of which had to be eliminated from the survey because they were improperly completed. The questionnaire was divided into three sections. The objective of the first section was to determine the respondents' in-home shopping habits, social class and stage in the family life-cycle. The second section was designed to test the respondents' personality and attitudinal char- éucteristics. The third section inquired as to the type of IDInoducts the respondents purchased by mail or by telephone, élrnd the respondents' activities during their leisure time. Data Analysis The data were analyzed in two steps. First, linear (i45-sscriminant analysis was used to test whether there were Eafirlzy significant differences between people who shOpped at l'1<:>1'ne and those who did not. Second, canonical analysis was ‘tl‘tuilized to determine if there were significant differences E‘ITuong individuals who purchased products through different tl’pes of in-home shopping outlets. Ezignear Discriminant Analysis Linear discriminant analysis permits the analyst to examine a set of independent variables to determine which, if any, are able to distinguish between two or more pre- determined dependent variables or classification boundaries. Statistically, the linear discriminant function maximizes 81 the independent variables.1 The discriminant function takes the following form: = + BX + Y aXl 2 cX3 The letters a, b and c represent the discriminant weights or coefficients, while X X represent the 1' X2' 3 numeric value which the respondents had on the respective variable. Y is the discriminant score which is computed for each respondent in the sample. There are two common techniques used to test whether the discriminant function significantly differentiated the respondent groups. The first test consists of classifying the subjects as belonging to one of the two respondent categories based upon their discriminant scores and the discriminant classification boundary. It is calculated by dividing the distance between the discriminant centroids in the two sample cases. The discriminant centroids are devel- oped by substituting the group means for each independent variable in the discriminant function and then solving for the Y value.2 That is, if the discriminant function was effective in distinguishing between the groups, it would be eXpected that a large percentage of the individuals in 1Franklin B. Evans, ”Psychological and Objective Factors in the Prediction of Brand Choice," Journal of Business, No. 32, October, 1959, pp. 340-369. 2William W. Cooley and Paul L. Lahnes, Multivariate Procedures for the Behavioral Sciences (John Wiley and Sons, Inc., 1962), p. 7. 82 Group I would have discriminant scores that would place them closer to the Group I centroid than to the Group II centroid; and a large percentage of the Group II respondents' dis- criminant scores would place them closer to the Group II centroid than to the Group I centroid. The first technique can be tested for statistical significance by using a chi square test to determine whether the percent correctly classified was greater than would be expected by chance. The second test of the power of the discriminant function is the Wilks' lambda criterion, which was developed by Rau in 1952,1 and determines the probability level for rejecting the null hypothesis incorrectly (Type I error).2 The results of both tests are reported in Chapter IV. The discriminant program3 used to test the data prints out three items which are not normally found in discriminant programs: (1) correlation of the independent variables with the discriminant scores, (2) the group means ,6f the independent variables, and (3) an F-test of the differences between the group means. By examining these items it is possible to estimate which of the independent 1Ibid., p. 7. 2Statistically, Wilks' takes the following form: = w/T where w is the pooled within—group deviations score cross products and T is the total deviation score cross-products matrix. Taken from Cooley and Lahnes, op. cit., p. 61. 3Donald J. Veldman, Fortran Programming for the Behavioral Sciences (New York: Holt, Rinehart and Winston, 1967), pp. 268-280. 83 variables are responsible for the effectiveness of the discriminant function in differentiating the two groups. That is, if a variable correlates highly (either positively or negatively) with the discriminant scores, then it can be assumed that this particular variable plays an important role in the discriminant function. In the same manner, if an independent variable's group means are not statistically significant from each other, then it can be assumed that the variables did not play an important role in the power of the discriminant function. Several statisticians have stated that an upward bias develOps in the predictive power of the discriminant function if the discriminant coefficients from a sample are used to compute the discriminant scores from the same sample.1 This problem can be attenuated by cross validating the data. To cross validate the data, the samples were divided in half. The discriminant program derives the discriminant coeffi- cients from the first half. These coefficients are then used to compute the discriminant scores and the discriminant centroids from the second half of the sample. In the research, the data was analyzed without cross- validating the data, and then the data was cross-validated 1J. W. Sheth, "Multivariate Analysis in Marketing," Journal of Advertising Research, February, 1970, pp. 24-39; and Ronald E. Frank, William F. Massy and Donald B. Morrison, "Bias in Multiple Discriminant Analysis,“ Journal of Market- ing Research, August, 1965, pp. 250-258. 84 to illustrate the differences in the two procedures. Also, any conclusions to be drawn from the data will be conserva- tive in nature. The results of both the discriminant analysis and the cross-validated discriminant analysis will be explained in Chapter IV. Canonical Analysis A second objective of the study was to determine if relatively distinct market segments could be located within the in-home market. Since in the research it was possible that a respondent purchase goods from any combination of several types of in-home shOpping outlets, canonical analysis was selected to test the data. Discriminant analysis could not be used to examine the second objective of the study, because of the underlying assumption of multiple discrim- inant analysis that the dependent variables be independent of each other. That is, a subject may only belong to one respondent group. As an example, with discriminant analysis, an individual would be classified as a department store in- home shOpper rather than as a department store in-home shopper and a specialty house in-home shopper. The respondents were asked, in questions #8 and #10 of Section I of the Questionnaire (see Appendix C), to indi- cate whether they purchased goods by mail or by telephone from large catalogue stores, novelty catalogue stores, club- of—the-month type of organization, specialty mail order houses, credit card organizations, or if they purchased 85 goods responding to advertisements in magazines and newspapers. The choice of these types of in-home selling organizations as a classification for the purposes of segmenting the in-home selling market, was based on obser- vation of the existing mail order and telephone market. The respondent could indicate that he purchased goods by mail or by telephone from one, from more than one, or from all the types of retailers listed in questions #8 and #10 of Section I of the Questionnaire. In order to verify if statistically significant groups of shOppers related to one or more groups of retail outlets could be obtained, canonical analysis was used. Mathematically, the objective of canonical analysis is to obtain a set of weights for each of the sets of vari- ables, such that the correlations between the weighted linear composites in each set are maximized.1 Whereas discriminant analysis maximizes the differences between two or more independent groups, canonical analysis maximizes the correlation between the criterion (or dependent) set of vari- ables and the predictor (or independent) set of variables. The interpretation of canonical analysis can be quite difficult. The common procedure is to examine each canonical root as an independent entity. The procedure 1Mark I. Alpert and Robert A. Peterson, "On the Interpretation of Canonical Analysis" (unpublished working paper 71-45, The University of Texas at Austin), 1971. 86 begins by examining the criterion and the predictor variables to determine which of the variables have high correlations with the respective roots.l No magic corre- lation is considered to be high, rather, the analyst should begin with the highest correlation and proceed until the 2 The researcher variables selected no longer "make sense." must be careful not to let his previously stated hypothesis influence the decisions. The canonical program develOps as many roots as there are criterion variables. Generally, though, only those roots which have levels of significance below some predetermined level, such as .05, are examined. It was thought that by examining the data with two different statistical techniques, and by dividing the data in a manner which would yield a core market and a broader market, it might be possible to learn more about the kinds of peOple who shop at home, those who do not, and the types of people who shop at various forms of in-home outlets. 1David L. Sparks and W. T. Tucker, "A Multivariate Analysis of Personality and Product Use," Journal of Marketing Research, February, 1971, p. 69. 2Alpert and Peterson, op. cit. 87 Summary In summary, Chapter III has presented the dependent and independent variables used in the research and the sample design. Also, the two multivariate statistical techniques used to determine if there are significant differences between those individuals who do and do not shop at home, and to test if market segments could be located among the in—home shoppers were presented. L/// CHAPTER IV PRESENTATION OF FINDINGS The objective of Chapter IV is to present the findings of the research. The chapter is divided into three segments. The first segment examines the difference between the active and the inactive in-home shOppers; the second segment examines the very active and the very inactive in- home shOppers. The third section reports the findings of the canonical analysis which was used to determine if dis- tinct market segments exist within the in-home shOpping market. Active vs. Inactive In-Home Shoppers Socioeconomic and Demographic Variables The first two research hypotheses dealt with socio- economic and demographic variables, and with the in-home shOpping patterns of the subjects. This portion of Chap- ter IV will examine these nine independent variables and their effects on in-home shopping behavior. Table 4 sum- marizes the socioeconomic characteristics, while Table 5 summarizes the demographic characteristics. 88 89 Hoo. mmmHo warp“: mmmao mappfle momma aawemu on» no mmmHo Hmfloom cofiumospo paocmmson vo. Esuuommm ascoflumospo mumsomuo umom one mo coon on» usonmsoucu oousnfluumflo cfl cofiumuucmocoo >>moz mo Ho>ma Hmcoflumoscm mo. ooo.m~w ooo.mmw uo>o pseudo av.ma um>o vacuum wm.Hm 6800c“ >Hflssm Hmuos moo. muoxuo3 mofl>umm paw .Hmofiuoao .mumom IcmE .Hmcofimmowoum mamfiowwwo 0cm muommcme paonmmson onu mo usozmsounu cousnfluumfio Hmowcnomu .Hmc0wmmowoum use: may no cowuwmsouo oocmofluficvflm muwmmocm muvmmonm owumauouomumnu mo Ho>mq osomucH,m>wuom:H oaomucH o>wuu¢ mmmmmozm mZO:qu m>HeumzH 02¢ m>HBO¢ mo mOHBmHmm90oHdEm m.m~ vo>onEm av.vm on» NO ucms>oamem cosmowuwcmflm muvmmonm mummmonm oeumfiuwuomumnu HMUfiumwumum mo Hm>wq waomIcH m>fiuomcH oEomIcH w>wuo¢ mmmmmomm mZOmIZH W>HBUma 00. um unmoww0cmwm .Hw>mH mo. um unmoflmficmflmo 0 .00>0H mo. um uGMUHchwflmn .Hm>0H OH. um ucmoflmflcmflmm 00.0 00.0- 000.0 00.0- 000.0 00.0: 000.0 00.0- 00000>000000 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00000000000000>00 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 000000 mmHSQEH 000300 opsufluud 00.0 00.0 00.0 00.0 000.0 00.0 00.0 00.0 000000 000300 00000000 00.0 00.0 00.0 00.0 00.0 00.0 000.0 00.0 000000000000000 00.0 00.0- 00.0 00.0- 00.0 00.0 00.0 00.0 000000 00 00000 000.0 00.0 000.0 00.0 000.0 00.0 000.0 00.0 00000 000000 00.0 00.0 00.0 00.0 000.0 00.0 00.0 00.0 000000000 000 00 0000 000 MO HO>0H HMGOwumodflm 00.0 00.0 000.0 00.0 00.0 00.0 00.0 00.0 000000 000000 00000 000.0 00.0 00.0 00.0 000.0 00.0 0000.0 00.0 000000000 000 00 @003 0:» mo cowummsooo 000.0 00.0 00.0 00.0 00.0 00.0 00.0 00.0 00000 0000 000000 00000 00000000000 00000 00000000000 00000 00000000000 00000 00000000000 lb COflUflHOHHOU lb COfiUMHOHHOU lh COHUMHQHHOU lb COHUMHOHHOU >H csm HHH cam HH :Dm H cam mmqm2¢m DmfidaHA¢> mmOfiU BZflQZflQMQZH mDOm mOh mmmmmOmm WZOEIZH H>HBU¢ZH 00000000> 024 m>HBU¢ mmB m0 Bzmummm ow mom Bmmfilm 024 mmOUw BZGZHZHmUMHQ WEB xBH3 mBZmHUHhmmOU ZOHBfiAmmMOU ma wHQMB 111 Very Active vs. Very Inactive In-Home Shoppers Socioeconomic and Demographic Variables The findings for the very active and the very inactive in-home shoppers will be reported in the same manner as the test results which were reported for the active and the inactive in-home shoppers. The findings begin with the socioeconomic and demographic variables which are summarized in Tables 20 and 21. 1. Occupation of the head of the household.--The very active in-home shoppers appear to have significantly higher status jobs than do the very inactive in-home shop— pers. As Table 22 indicates, 42 percent of the "very active in-home shOppers" in the sample, have "professional," or "technical" positions, while 21.6 percent of the very in- active shoppers fall in this category. A second category which exemplifies the differences in the two groups is the "service workers." In that category are 5.7 percent of the very active and 18.2 percent of the very inactive in-home shoppers. The differences are statistically significant at the .001 level with a two-tailed test. 112 unwofiwflcmflm 002 HO. uc00000co0m uoz uc00000cq0m uoz #:00000cm00 uoz cumuumm 00:00000 oz 00000 00 mom 00 00000 00 00.00 00000 om um>o 000 0000008 0003 00.00 mucmpcommwu 0000008 >00H08000 0000 0000 00000000 00.00 cumuumm uocHume oz 00000 00 0mm 00 00000 00 00.00 00000 om uw>o 000 0000008 0003 00.00 mucmpcommmu 0000008 >HHH08000 0000 0000 00000000 00.00 0000010000 000800 000000000 000 00 0000 000 00 000 0000008 0000» 00 000852 050000 000000: 00008080: 0:» mo u:08>oam8m 00:00000cm0m HmowumHumum mo 00>00 muwmmonm wEOmch m>wuomcH >um> mummmonm wEomucH m>0004 >Hm> 00um00000000£0 mfimmmomm mZOmIZH mmmmmomm mZOmIZH m>HBU< >mm> m0 MUHBMHmmevdmde UHmmdmOOZmo W>HBUdZH >mm> 02¢ om manna 113 000. 00000 000002 00000 000008 300m 00000 000000 00000003000 uoz 000050000 000050000 000300 000300 3003 000 0000000 00050000 000 0000000 00000050m mo. ooo.mm0 00>0 00.00 ooo~me 00>0 0mm 080000 0000000000 000. QOn 030 0503050030 000x003 0000:3000 :00050000000 30>0 000000 030 000000000000 :00u0m5000 003000000000 00000030 000mm030 00000000000030 00000000000 00 00>00 080mIcH 0>000000 >00> 0800100 0>000< >00> mdmmmomm mZOmIZH m>HBUmm> 02¢ mmmmmomm MZOmIZH m>HBU( >mm> m0 mUHBmHmm804m¢IU UHZOZOUNOHUOM 0m 00308 114 Table 22 OCCUPATION OF THE HEAD OF THE HOUSEHOLD OF VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Very Active Very Inactive Occupations In-Home Shoppers In-Home Shoppers 2 1 p. 1 Professional, technical and kindred workers (91-100) 37 42.0 19 21.6 Managers, officials, and proprietors except farm (81-90) 26 29.5 25 28.4 Clerical, sales, and kindred workers (71-80) 9 10.2 15 17.0 Craftsman, foreman, and kindred workers (61-70) 8 9.1 6 6.8 Operatives and kindred workers (51-60) -- -- 2 2.3 Semiskilled labor (41-50) 3 3.4 5 5.7 Service workers (31-40) 5 5.7 16 18.2 Unskilled labor (21-30) —- -— _- __ Babysitters (less than 20) -- -- -- -- Total 88 99.9 88 100.0 ‘/ 115 2. Total family_income.--Table 23 shows that there is an indication that the very inactive in-home shoppers tend to have higher total family income levels than do the very inactive in-home shoppers. Twenty-five percent of the very active in-home shoppers in the sample had family incomes in excess of $25,000, while 11.5 percent of the very inactive in-home shoppers had income levels of over $25,000. differences in income levels were significant at the .03 level with a two-tailed test. Table 23 INCOME LEVELS OF VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Income Very Active In-Home Shoppers Very Inactive In-Home Shoppers Less than $5,000 $5,000-$7,999 $8,000-$9,999 $10,000-$l4,999 $15,000-$24,999 $25,000 and over Total 2_ 3 4 9 25 25 22 88 '09 fl_ 4 11 13 23 26 10 87 [a 12.6 14.9 26.4 29.8 11.5 99.8 116 3. Level of education of the head of the household.-- The heads of the households in the very active in-home shop- per categories appear to have a higher level of education than do heads of the households in the very inactive in-home shopper families. However, as Table 24 indicates the differ- ences were not strong and the data were not statistically significant. Table 24 EDUCATION OF THE HEAD OF THE HOUSEHOLD OF VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Very Active Very Inactive Education In-Home Shoppers In-Home Shoppers I: lw I: |m Elementary one and two years -- —— _- __ three and four years -— -- _- -_ five and six years -- —- -_ -- seven years -- -- _- -_ eight years -- -- 4 4.6 High School one year 1 1 3 two years 4 4. 2 three years -- -- 1 four years 15 17.0 26 College one year 4 two years 8 three years 3 four years 21 Graduate School 32 36.4 15 17.2 Total 88 99.9 87 99.8 117 4. Social c1ass.--The very active in-home shoppers appeared to belong to a significantly higher social class than did the very inactive in—home shoppers. Table 25 shows that 80.7 percent of the very active in-home shoppers and 48.9 percent of the very inactive in-home shoppers in the sample, scored between 81—100 on the social class index. In contrast, 39.7 percent of the very inactive in-home shOppers and 14.8 percent of the very active in-home shop- pers scored between 61-80 on the social class index. The differences were statistically significant at the .001 level. Table 25 SOCIAL CLASS OF VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Very Active Very Inactive Gradations In-Home Shoppers In-Home Shoppers 2. 3. a 3.: 91-100 49 55.7 32 36.4 81-90 22 25.0 11 12.5 71-80 8 9.1 23 26.1 61-70 5 5.7 12 13.6 51-60 2 2.3 6 6.8 41-50 1 1.1 2 2.3 31-40 1 1.1 2 2.3 Less than 30 -- -_ -- -_ Total 88 100.0 88 100.0 118 5. Employment of the homemaker.-—Table 26 shows that 37.5 percent of the very inactive in-home shoppers in the sample were employed full time; and 31.8 percent of the very inactive in-home shoppers were employed full time; 52.3 percent of the former group and 56.8 percent of the latter group were unemployed. The differences in the data were not statistically significant. Table 26 EMPLOYMENT OF THE HOMEMAKER OF VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Employment Very Active Very Inactive Status In-Home Shoppers In-Home Shoppers 2 ‘3; 2 3; Unemployed 46 52.3 50 56.8 Part-time 9 10.2 10 11.4 Full-time 33 37.5 28 31.8 Total 88 100.0 88 100.0 119 6. Marital status.--More than 85 percent of both respondent groups were married. There were no statistically significant differences between the two in-home shopping groups. The data are reported in Table 27. Table 27 MARITAL STATUS OF VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Very Active Very Inactive Marital Status In-Home Shoppers In—Home Shoppers 2 i 2 3: Married 75 85.2 77 87.5 Single _13| 14.7 ._11 12.5 Total 88 99.9 88 100.0 v//;. Number of years married.--The very inactive in- home shoppers tended to be married for a longer time than did the very active in-home shoppers. As Table 28 indicates, 13 percent of the very active group and 5.2 percent of the very inactive group had been married for less than 8 years. Also, 50.6 percent of the very active in-home shoppers and 71.4 percent of the very inactive in-home shoppers have been married for over 20 years. The differences were not statis- tically significant. 120 Table 28 NUMBER OF YEARS MARRIED OF VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Very Active Very Inactive Years Married In-Home Shoppers In-Home Shoppers 1 1 1 1 Less than 3 years 4 5.2 -- -- 4-7 years 6 7.8 4 5.2 8-12 years 14 18.2 5 6.5 13-20 years 14 18.2 13 16.9 Over 20 years _22_ 50.6 _§§| 71.4 Total 77 100.0 77 100.0 »//8. Age of the head of the household.--The heads of the households in the very inactive in-home shopper cate- gories appeared to be older than the heads of the household in the very active in-home shOpper categories. Table 29 shows that 13.7 percent of the very active category and 28.69 percent of the very inactive group had heads of the households who were 60 years of age or older. In contrast, 16.1 percent of heads of the households of the very active in-home shOppers were younger than 35 years of age compared to 10.2 percent for the very inactive in-home shoppers. The differences were statistically significant at the .01 level for a two tailed test. 121 Table 29 AGE OF THE HEAD OF THE HOUSEHOLD OF VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Chronological Very Active Very Inactive Age In-Home Shoppers In-Home Shoppers 2 .1 .2 .1 18-20 1 1.2 -- -- 21-24 -- -- —- —- 25-29 4 4.6 5 5.7 30-34 9 10.3 4 4.5 35-39 14 16.1 6 6.8 40-44 .‘ 11 12.6 8 9.1 45-49 12 13.8 13 14.8 50-54 15 17.2 10 11.3 55-59 9 10.3 8 9.1 60-64 9 10.3 19 21.6 70 and over .__3 3 4 15 17.0 Total 87 99.8 88 99.9 122 9. Family life—cycle.--No distinct patterns were found between the two respondent groups for the variable family life-cycle. The data are presented in Table 30 and were not statistically significant. Table 30 FAMILY LIFE-CYCLE OF VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Stage in the Family Very Active Very Inactive Life-Cycle In—Home Shoppers In-Home Shoppers .11 31 9. 35. Single 8 9.1 8 9.1 Couple without children 20 22.7 34 38.6 Oldest child three years old or younger 2 2.3 l 1.1 Oldest child four to five years of age 5 5.7 2 2.3 Oldest child six to thirteen years of age 17 19.3 10 11.3 Oldest child fourteen to nineteen years of age 26 29.5 21 23.9 Oldest child over twenty years of age with brother or sisters younger than twenty 8 9.1 6 6.8 Youngest child over twenty years of age 2 2.3 6 6.8 Total 88 100.0 88 99.9 123 Sociopsyghological Variables Of the six sociopsychological variables used in the research only "cosmopolitanism" and "conservatism" differen— tiated significantly between the very active and the very inactive in-home shoppers analyzed in the sample. The former group appeared to be more liberal and more cosmo- politan than the latter group. The differences were significant at the .03 and .01 levels, respectively. Discriminant Analysis The same eleven variables which were used to differentiate active and inactive in—home shoppers were used to differentiate very active from very inactive shoppers. The data were analyzed first without cross validation and then with cross validation. The results of both procedures are reported below. The Wilk's lambda test of the non-cross validated data indicates that the discriminant function was able to significantly differentiate between the two respondent groups; Wilk's lambda had a value of 0.835 which is sta- tistically significant at the .002 level. The discriminant classification matrix which is reported in Table 31 also indicates that the two respondent groups are significantly different from each other; 51 very inactive in-home shoppers were classified correctly and 37 were classified incorrectly. Sixty-five very active in-home 124 Table 31 DISCRIMINANT CLASSIFICATION MATRIX FOR VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Very Inactive Very Active In-Home Shoppers In-Home Shoppers Very inactive in-home shoppers 51 37 Very active in-home shoppers 23 65 shoppers were classified correctly and twenty-three were classified incorrectly. The chi square test of statistical significance of the two groups had a value of 16.99 which is significant at the .001 level. Table 32 lists the correlation coefficients and the F-tests for the differences in group means for the two respondent groups. It is interesting to note that the vari- ables which had high correlations with the discriminant scores and significantly different group means for the active and inactive in-home shoppers also tended to have high correlations and significantly different group means. As in the case with the active and inactive in-home shoppers, a cross validation of the data reduced the signif- icance of the findings. Table 33 shows that Wilks' lambda value was significant at the .05 level for only the first run, while the chi square test of significance of the 125 Table 32 CORRELATION COEFFICIENTS WITH THE DISCRIMINANT SCORE AND F-TEST FOR THE DIFFERENCES IN GROUP MEANS FOR VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Correlation Variable Coefficients F-Ratio Family life cycle 0.22 1.40 Occupation of the head of the b household 0.64 12.65 Total family income 0.41 4.96a Educational level of the head of the household 0.28 2.21 . b Soc1al class 0.69 14.97 Trust in people -0.13 0.51 Cosmopolitanism 0.50 7.57b Attitude toward credit 0.21 1.29 Attitude toward impulse buying -0.06 0.09 Adventuresomeness 0.23 1.64 Conservatism —0.39 4.59a aSignificant at .05 level. bSignificant at .01 level. 126 Table 33 WILK'S LAMBDA AND CHI SQUARE TEST FOR STATISTICALLY SIGNIFICANT DIFFERENCES ACHIEVED BY THE THREE INDEPENDENT CROSS VALIDATED RUNS ON THE VERY ACTIVE AND VERY INACTIVE IN-HOME SHOPPERS Run Wilk's Lambda Chi Square 1 0.816C Not significant 2 0.756d Not significant 3 0.855a Not significant aSignificant at the .20 level. bSignificant at the .10 level. CSignificant at the .05 level. dSignificant at the .01 level. discriminant classification matrices was not significant at the .05 level for any of the three runs. The correlation coefficients with the discriminant scores and the F-test for the differences in group means for the very active and very inactive in-home shoppers are located in Table 34. 127 .Hw>mH Ho. um ucmoflmwcmfim U n .Hm>wa mo. um ucmofimacmwmo .Hm>ma oa. um u:MOHMHcmnmm mmb.m oa.o No.0 no.0 hv.m MN.H No.5 vm.o mmv.m oov.m H©.o mv.ou mo.OI no.0! mo.o ov.o mN.OI 00.0 ma.o 5v.o mm.o ON.O .Hm>mH mo. um uanAMAcmnm 6H.o no.0 mm.a m~.o sm.H mm.o mma.m 4m.o ovm.m 64.0 64m OH oo o- 6m.o NH.o 64.0 4H.o mm.o HH.o- Hm.~ mm.o mm.o mH.o om.o ha.o HN.O NN.N Hm.v ovm.m ¢O.V NH.o no.0 Dom m om.o ON.OI mo.o OH.O mm.o ov.o vm.on mv.o mo.o wH.o Hm.o HN.O EWHUM>H0mCOU mmOCOEOmGHSUCO>U4 mcflhsn mmasmefl @Hm3ou modufluud unomuo amazon mwsufluua EmflcmuHHOQOEmoo mamomm CH undue mmMHo HmHoom muonmmsos mg» no name msu mo Hm>ma Hmcoflumospm mEOUCH uafismm Hmuoe caonwmson 0:» mo cam: on» NO cowummsooo oao>o mmaa >HHEmm oflumm-m ucmfiufiuumoo coHumHmuuoo oflummlm ucmfloflmmwoo coAumHmuuou HH Cflm enummum ucmuoflmmmoo COHUMHOHHOU H Cfim moanwfium> mWAAde zoozfim DWB€0H4¢> memU BzmazmmmozH mummfi mom mmmmmomm mZOmIZH m>HBU¢ZH >mm> 024 W>HBU¢ Nmm> mmB vm OHQMB m0 Bzmummm ow mom Emmelm 92¢ mmOUw BZdZHZHmUmHD mmfi mEHZ wBZmHUHmhmOU ZOHBdAmmmOU 128 Canonical Analysis Canonical analysis was used to analyze the active in-home shoppers and the very active in-home shOppers. It was felt that this particular multivariate statistical technique might be successful in locating market segments within the in-home shOpping market. Table 35 shows the results of the canonical analysis of the active in-home shoppers. Since it is difficult to interpret the findings of canonical analysis beyond the significant roots, only the first three roots are presented in this table. In general, with canonical analysis, if there are P predictor variables and C criterion variables, and P >C, there will be C pairs of canonical functions calculated. The functions are independent from each other except for their pairmates in the remaining set of variables to which they are maximally correlated. The functions are calculated in descending order from the residual variance of the pre- vious functions. The roots, or squared canonical coeffi- cients are analogous to the index of determination in regression analysis and represent the amount of shared variance explained by the variables in each function.1 The first three roots had canonical R of 0.402, 0.348, and 0.324 which were significant at the 0.004, 0.032 and 0.039 levels, respectively. The fourth root had a 1Alpert and Peterson, op. cit., p. 3. 129 Table 35 RESULTS OF THE CANONICAL ANALYSIS FOR ACTIVE IN-HOME SHOPPERS Canonical Coefficients Variables l 2 3 Predictor Set Family life-cycle -0.2187 -0.5361 -0.l406 Occupation of the head of the household 0.5503 -0.0024 0.4254 Total family income 0.3504 0.1064 0.5502 Education level of the head of the household 0.6502 -0.2880 0.2426 Social class 0.6214 -0.0505 0.4845 Trust in people -0.0669 —0.4400 0.2410 Cosmopolitan attitudes 0.2656 -0.4260 0.0853 Attitude toward credit -0.0276 0.1406 0.2958 Attitude toward impulse buying 0.2761 0.0448 0.1319 Adventuresomeness 0.4687 —0.1360 -0.4359 Conservatism -0.2537 0.1694 -0.2779 Criterion Set Large catalogue outlets 0.5400 0.1049 0.0231 Novelty outlets 0.3337 0.7615 0.2442 Club of the month 0.4035 0.0287 0.0218 Specialty outlets 0.6135 -0.3511 0.4749 Credit card sales -0.3147 0.2057 0.7028 Magazine purchases 0.2285 -0.1014 0.0616 Other direct mail 0.2262 -0.0590 -0.l3l9 Roots 0.1614 0.1210 0.1054 Canonical R 0.402 0.348 0.324 x2 36.693 26.665 23.327 D.F. 17.0 15.0 13.0 Probability 0.0044 0.0329 0.0392 130 canonical R of 0.278 and was not statistically significant (0.113 level). The above levels of significance of the first three roots indicate there clearly are significant relationships between the criterion and the predictor set of variables. The first root is associated with large catalogue shOppers, club-of-the-month shoppers and specialty shoppers. These individuals can be best described as having high status occupations, high income levels, high levels of education, high social class and adventuresome attitudes. It appears that individuals who are of relatively high social class and are somewhat adventuresome tend to shop at the fairly well established in-home shopping outlets of large catalogue stores, club-of-the-month organizations and specialty houses. The second root is associated with individuals who utilize novelty in-home shopping outlets and who do not use specialty organizations. These two groups tend to be in the early stages of the family life cycle, tend not to trust people and are not very cosmopolitan in their orientation. The third root is associated with individuals who shop via specialty and credit card in-home shopping outlets. The predictor set of variables which is related to the two groups are high status occupations, high total family in- come, high social class and non-adventuresome attitudes. 131 The predictor set in the third root is similar to the predictor set in the first root, except that the adventure- someness variable is reversed. Since the criterion set of variables are substantially different between the first and the third root, then it can be assumed that the key predic- tor variable which is distinguishing between the first and the third root is the adventuresomeness. The results of the canonical analysis on the very active in-home shoppers was not as clear as it was on the active in-home shoppers. Only the first root was statis- tically significant; it had a canonical R of 0.567 and significance level of 0.015. As Table 36 shows this factor was associated with individuals who do not shop at large catalogue outlets or specialty outlets, but who do shop via credit card promotion and other direct mail sources. The predictor set which related with these four groups were: early stages in the family life cycle, low income levels, low social status, and conservative attitudes. Since only one root was significant, there is no way that these results can be generalized. 132 Table 36 RESULTS OF THE CANONICAL ANALYSIS FOR THE VERY ACTIVE IN-HOME SHOPPERS Canonical Coefficients Variables 1 — Predictor Set Family life cycle . . . . . . . . . . . . . . . -0.4020 Occupation of the head of the household . . . . -0.239l Total family income . . . . . . . . . . . . . . -0.3867 Educational level of the head of the household . . . . . . . . . . . . . . . . . . -0.0524 Social class . . . . . . . . . . . . . . . . . . -0.3430 Trust in people . . . . . . . . . . . . . . . . —0.2034 Cosmopolitan attitudes . . . . . . . . . . . . . -0.1552 Attitude toward credit . . . . . . . . . . . . . 0.2925 Attitude toward impulse buying . . . . . . . . . 0.1349 Adventuresomeness . . . . . . . . . . . . . . . -0.1570 Conservatism . . . . . . . . . . . . . . . . . . 0.3644 Criterion Set Large catalogue outlets . . . . . . . . . . . . -0.4090 Novelty outlets . . . . . . . . . . . . . . . . 0.1181 Club of the month . . . . . . . . . . . . . . . 0.2731 Specialty outlets . . . . . . . . . . . . . . . -0.5172 Credit card sales . . . . . . . . . . . . . . . 0.5158 Magazine purchases . . . . . . . . . . . . . . . -0.169l other direct mail . O O O C O O I O O O O O O . Roots . . . . . . . . . . . . . . . . . . . . . . 0.3391 Canonical R7.. . . . . . . . . . . . . . . . . . . 0.567 X2 . . . . . . . . . . . . . . . . . . . . . . . . 32.1020 D.F. . . . . . . . . . . . . . . . . . . . . . . . 17.0 Probability . . . . . . . . . . . . . . . . . . . 0.0157 133 Conclusion Chapter IV has presented the findings of the study. Although the result of most empirical studies are never as clear cut as the researcher would like, it should be apparent that through the use of discriminant analysis distinctions have been located between both in-home shoppers and individuals who do not shop at home, and between the various types of in-home shoppers. Chapter V will evaluate the research hypotheses in view of the findings. Major conclusions will be drawn also in that chapter, from the research findings in view of its objectives. CHAPTER V EVALUATION OF THE RESEARCH HYPOTHESES The objective of Chapter V is to compare the findings of the research with the research hypotheses. The chapter is divided into three sections. In the first section an evaluation is made of the selection of the independent variables to measure the validity of the research hypotheses and in relation to the research objec- tives. The second section evaluates the research hypotheses based on the findings presented in Chapter IV. The last section analyzes the major conclusions that can be drawn from the research findings in view of its objectives. Evaluation of the Selection of the Independent Variables The major problem found by the researcher, when attempting to design a study that would provide answers to the first objective of the research, i.e., to determine whether or not people who shop at home can be differentiated from those who do not, was that of selecting the independent variables. It was necessary to determine which characteris- tics should be analyzed in order to provide a meaningful 134 135 profile of the in-home shopper. It seemed necessary to start with socioeconomic and demographic variables in that these have proven in part to be effective tools in market segmentation. The choice of the sociopsychological variables was made by personal judgment, based upon opinions of writers as to what might cause people to purchase goods by mail or by telephone. Indications were found in the literature which suggested that the colorful pages of mail order cata- logues create a desire to buy.1 A conclusion could be drawn from such a statement as to the attitudinal characteristics of in-home shoppers; it was assumed that in-home shoppers would have a stronger tendency to impulse buying than individuals who do not shop at home. Other references in the literature are made to the fact that the customer of in-home retailers would purchase from them partially because of the fact that the items sold by retailers are not available from local stores.2 This statement suggests the possibility that in-home shoppers might be more cosmopolitan than people who do not shop at home. Therefore, a measure of cosmopolitanism was l"More Firms Seen Selling Direct to the Consumer," Home Furnishing Daily, May 1, 1970, p. 7. 2Charles F. Higgins, "The Booming In-Home Market," The Reporter of Direct Mail Advertising, Editor's Choice, Summer, 1967, pp. 48-49. 136 introduced as an independent variable to be tested in the research. The use of credit cards as well as other forms of credit has been mentioned in the literature as one of the possible causes for the growth of several forms on in-home retailing.1 Therefore, one of the independent variables introduced in the research related to the subjects attitude toward credit. In an effort to segment customers, reference in the literature was found to the fact that those who could be classified as "creative consumers" were considered to be more adventurous in their choice of products and less con- servative in their purchases and investments.2 It was felt that both variables, adventuresomeness and conservatism might be valuable instruments in differentiating in-home shoppers from people who do not purchase at home, therefore, they were included in the study. Consumers' perception of risk has been analyzed in relation to their purchasing behavior and especially with 3 reference to mail order sales. A parallel was assumed to 1Lester Wundermann, "Mail Order, The Coming Revolu- tion in Marketing," Advertising Age, July 24, 1967. 2Emanuel Demby and Louis Cohen, "Consumer Life Style More Vital to Marketing Than Income Level," The Reporter of Direct Mail Advertising, June, 1969. 3Maxwell Sroge, "The Distribution Revolution . . . Anatomy of Direct Marketing," The Reporter of Direct Mail Advertising, July, 1970, pp. 18-19. 137 exist between individuals' perception of risk and their trust in peOple. In order to verify if respondents' trust in people could differentiate in-home shoppers from not in- home shOppers, a measure of such attitude was introduced in the research as an independent variable. Evaluation of the Research Hypotheses The primary scope of the evaluation of the research hypotheses is to allow the research to draw broader implica- tions from the analysis by integrating the findings with the objectives of the study. The hypotheses and subhypotheses, enumerated in Chapter I of the study, are presented along with the most significant conclusions which can be drawn from the research findings. Social Class I. Ceteris Paribus, among all respondents there will be no significant differences between people who shop at home and people who do not shop at home, when classified by their respective social classes. On the basis of the results presented in Tables 9 and 25, there seem to be significant differences between people who shop at home and people who do not shop at home, when classified according to their social class. Table 25 particularly indicates that very active in-home shoppers when compared with very inactive in-home shOppers scored 138 significantly higher in social class. The results, therefore would tend to indicate that the alternative hypothesis should be accepted. Occupation I. A. Ceteris paribus, among all respondents, there will be no significant differences between in-home shop— pers and people who do not shop at home, when clas- sified by their members' respective occupation. On the basis of the findings presented in Tables 6 and 22, the in-home shOppers analyzed in the sample were found to be primarily in the two top job categories. Table 22 shows a comparison between the very active and the very inactive in-home shoppers in terms of the occupation of the head of the household. The difference between these two groups appears to be even greater than that existing between active and inactive in-home shoppers. Families headed by professionals or technical workers are more likely to be concentrated in the very active in—home shoppers group, while families headed by service workers are probably con- centrated in the very inactive in-home shoppers group. On the basis of the findings, therefore, there seem to be significant differences between in-home shoppers and peOple who do not shop at home, when classified by the occupation of the head of the household. -'/ 139 Income Level ,I. B. Ceteris paribus, among all respondents, there will be no significant difference between in—home shop- pers and people who do not shop at home, when classified by their respective income level. The results presented in Tables 7 and 23 illustrate the relationship between the in—home shopping activity and the income level of the respondents. The results appear to indicate that in-home shOppers can be differentiated from people who do not shop at home, when classified by their respective income level. When active in-home shoppers were compared to inactive in-home shoppers, the active in-home shoppers appeared to concentrate more in the over $25,000 category, as shown in Table 7. The difference between very active in-home shoppers and very inactive in-home shoppers seemed to be even greater. While 25 percent of the very active in—home shOppers classi- fied in the over $25,000 category, only 11-1/2 percent of the very inactive in-home shoppers were found to be in the same category as shown by Table 23. Educational Level of the Head of the Household I. C. Ceteris paribus, among all respondents, there will be no significant difference between in-home shop- pers and people who do not shop at home, when classified by their members' respective level of education. 140 As indicated by the results in Tables 8 and 24, the in-home shoppers could probably be differentiated from people who do not shop at home, by the level of education of the head of the household. Table 8 indicated that families headed by individuals who attended graduate school tended to be more heavily concentrated with the active in- home shoppers. Similar results were found when the very active in-home shOppers were compared to the very inactive in-home shoppers. Table 24 indicates that heads of the households of the very active in-home shoppers category are more likely to have a higher level of education than heads of the households in the very inactive in-home shoppers category. Family Life-Cycle II. Ceteris paribus, among all respondents, there will be no significant difference between peOple who shop at home and people who do not, when classified by their stage in the family life-cycle. On the basis of the data presented in Tables 14 and 30, this hypothesis must be accepted, i.e., the research data appears not to support the conclusion that a relation- ship exists between stage in the family life-cycle and the in-home shopping patterns of individuals. Although the analysis of research results presented in Table 14 shows that the active in-home shoppers tend to be slightly "older" 141 in their stage in the family life-cycle, than inactive in-home shoppers, such a difference was not found between the very active and the very inactive in-home shoppers. Therefore, no difference is likely to exist between in-home shoppers and not in-home shoppers when classified by their stage in the family life-cycle. Marital Status II. A. Ceteris paribus, among all respondents, there will be no significant difference between in-home shop- pers and people who do not shop at home, when classified by their respective marital status. On the basis of the results presented in Table 11, no statistically significant difference was found between the active and the inactive in-home shoppers in the sample when compared as to their marital status. Similar results were obtained when the same variable was used to compare the very active and the very inactive in-home shOppers (Table 27). It seems unlikely, therefore, that in-home shoppers can be differentiated from not in-home shoppers as to their marital status. Age of the Head of the Household II. B. Ceteris paribus, among all respondents, there will be no significant difference between in-home shop- pers and people who do not shop at home, when classified by the age of the head of the household. 142 A comparison of the active and inactive in-home shoppers based on the age of the household head is shown in Table 13. No significant difference appeared to exist between active and inactive in-home shoppers as to the age of the household head. Table 29, however, presents different results. A comparison made between the very active and the very inac- tive in-home shoppers as to the age of their household head showed that in the very active in-home shoppers' category, the household heads were younger than those of the very inactive in-home shoppers category. There seems to be some correlation, therefore, between the frequency of in-home shopping activities and the age of the household head. The evidence, however, is not sufficient to reject the above hypothesis. Employment Status of the Homemaker II. C. Ceteris paribus, among all respondents, there will be no significant difference between in-home shop- pers and people who do not shop at home, when classified by the employment status of the homemaker. Tables 10 and 25 show that there appears to exist a relationship between the employment status of the homemaker and the in-home shOpping activity. A greater percentage of the active in-home shoppers were employed full time, as compared to the inactive in—home shoppers. The results, “-9“ w. 1' 143 however, were not statistically significant. It seems unlikely, therefore, that in-home shoppers can be differ- entiated from not in-home shoppers, based on the employment status of the homemaker. Number of Years Married II. D. Ceteris paribus, among all respondents, there will be no significant difference between in-home shop- pers and people who do not shop at home, when classified by the number of years they were married. On the basis of the findings presented in Tables 12 and 28, the above hypothesis appears to be validated with respect to the relationship between the number of years the respondents were married and their in-home shopping activ- ities. The difference found between the very active and the very inactive in-home shoppers in the sample as to the number of years married was. not significant, although it seemed that the very inactive in-home shoppers tended to be married for a longer time. Adventuresomeness III. Ceteris paribus, among all respondents, there will be no significant difference between people who shop at home and people who do not shop at home, when classified by their respective adventuresomeness. 144 On the basis of data presented in Tables 15 and 32, the above hypothesis seems to be validated. When active and inactive in-home shoppers were compared as to their adventuresomeness, no distinct pattern was found which differentiated among them. The same results were obtained when a comparison was made between the very active and the very inactive in-home shOppers. Therefore, there seems to be no relationship between the respondents' attitude toward adventuresomeness and their in-home shopping habits. Cosmopolitanism IV. Ceteris paribus, among all respondents, there will be no significant difference between people who do shop at home and people who do not shop at home, when classified by their respective degree of cosmopolitanism. As Table 15 shows, when active and inactive in-home shoppers' cosmopolitan attitudes were compared, the active in-home shoppers tended to display a more cosmopolitan orientation than did the inactive in-home shoppers. Similar results were obtained through a comparison of very active and very inactive in—home shoppers, as to their attitudes toward cosmopolitanism. As the results of this test were statistically significant both times, the above hypothesis should probably be rejected. That is, people who shop at home probably can 145 be differentiated from those who do not, when compared as to their cosmOpOlitan attitudes. Trust in People V. Ceteris paribus, among all respondents, there will be no significant difference between people who shop at home and people who do not shop at home, when classified by their respective trust in people. The results shown by Tables 15 and 32 indicate that no relationship appeared to exist between in-home shOpping habits of the respondents and their trust in people. When active and inactive in-home shOppers were compared as to their trust in people, no distinct pattern was found that could indicate differentiation between the two groups. Similar results appeared when the very active and the very inactive in-home shoppers' trust in people was measured; no strong relationship seemed to exist between the above atti- tude and the respondents in-home shopping activities. The above hypothesis, therefore, is likely to be validated in that there seem to be a relationship between trust in peOple and the in-home shopping patterns of the sample. Conservatism VI. Ceteris paribus, among all respondents, there will be no significant difference between people who shop at home and people who do not shop at home, !——v wv‘n 146 when classified by their respective conservative- liberal attitudes. Table 15 shows that when active and inactive in-home shoppers were compared as to their conservative-liberal attitudes, the active in-home shoppers showed that they tended to be more liberal than the inactive in—home shoppers. Table 32 shows similar results for the comparison made r between very active and very inactive in-home shoppers. The results were statistically significant for both groups. It appears, therefore, that in-home shoppers can I” 1 be differentiated from people who do not shOp at home, when compared as to their conservative-liberal attitudes. Attitude Toward Credit VII. Ceteris paribus, among all respondents, there will be no significant difference between people who shop at home and people who do not shop at home, when classified by their respective attitude toward credit. Tables 15 and 32 show that the in-home shoppers in the sample tended to have a more favorable attitude toward credit than peOple who did not shop at home. Both compar- ison of active and inactive in-home shOppers, as that of the very active and inactive in-home shoppers indicate that there is evidence contrary to the above hypothesis. It appears, from the results, that in-home shoppers can be 147 ciigfferentiated from peOple who do not shop at home by tliezir attitude toward credit. Ikerzdency for Impulse Buying ‘VJCII. Ceteris paribus, among all respondents, there will be no significant difference between people who shop at home and people who do not shop at home, when classified by their respective tendency for impulse buying. As appears in Tables 15 and 32, when the active and iJiactive in-home shoppers in the sample were compared as to tflieir tendency for impulse buying, no distinct pattern was kaund, and therefore, no differentiation could be made on the basis of this particular variable. The same results ‘Nere attained when very active and very inactive in-home Shoppers were compared. No strong relationship was found between the respondents' tendency to impulse buying, and their in-home shOpping habits. In view of these results, the above hypothesis should probably be accepted as to the relationship between in-home shopping, and people's attitude toward impulse buying .~~-. , v.-‘——. v‘r - u 1— -1 - ‘. M_—\—I'"Y 148 Evaluation of the Second Set of Hypotheses The second set of hypotheses stated in Chapter I relates to the comparison between different segments of in-home shoppers. A summary of the hypotheses, as well as of the findings related to them will be explained below. Three groups of in-home shoppers, or three sub— markets, resulted from the analysis of the second set of hypotheses. In the first group were classified those in- home shoppers who purchased goods from large catalogue stores, from club-of-the-month type of organizations, and from specialty in-home retailers. The second group was formed by in-home shoppers who purchase goods from novelty in-home retailers, and who do not purchase from specialty organizations. The third group is made up of those indi- viduals who shop at home from specialty retailers, and from credit-card in-home outlets. Each of these groups will be compared as to the validity of the hypotheses regarding market segmentation within the in-home shopping market. First Group This group consists of in-home shoppers who purchase from large catalogue stores, club-of-the-month organizations, and in-home retailers. 149 Social Class I. Ceteris paribus, among all in-home shoppers, there will be no significant difference, when classified by their respective social class. As can be seen from the results in Table 35, the in-home shoppers classified in the first group were found to appear to have high social class. It seems, therefore, that in-home shoppers who purchase their goods from large catalog stores, club-of-the-month organizations, and specialty retailers, can be differentiated from all other in-home shOppers, when classified by their respective social class. Occupation I. A. Ceteris paribus, among all in-home shoppers, there will be no significant difference, when classified by their respective occupation. When the in-home shoppers classified in the first group were compared to all other in-home shoppers in the sample, it was found that they tended to have a higher status occupation than the latter. For this reason, the alternative hypothesis can probably be accepted, i.e., in- home shoppers who purchase their products from large cata- logue stores, club-of-the-month organizations, and specialty retailers, can be differentiated from all other in—home shoppers, when classified as to their occupation. -" '.-|-‘i lawn-n 150 Income Level I. B. Ceteris paribus, among all in-home shoppers, there will be no significant difference, when classified by their respective income level. The results shown in Table 35 indicate that in-home shoppers pertaining to the first group tend to have a higher income level than all other in-home shoppers. That is, income level appears to be useful in differentiating among segments of in-home shoppers. Educational Level I. C. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective educational level. This hypothesis, as those previously considered, was rejected in terms of the findings shown in Table 35. In fact, when in-home shoppers who purchase their goods from large catalogue stores, club-of-the-month organizations and specialty retailers were compared to all other in-home shoppers, they indicated to have a higher educational level than the latter. Therefore, educational level is a variable which can be used to discriminate among in-home shoppers. Family Life-Cycle II. Ceteris paribus, among all in-home shoppers, there will be no significant differences when classified by their respective stage in the family life-cycle. 151 The results in Table 35 indicate that this hypothesis is valid with respect to in-home shoppers per- taining to the first group. In fact, no difference was found among in-home shoppers who purchase products from large catalogue stores, club-of—the-month organizations and from specialty retailers, and all other in-home shOp- pers, when classified by their stage in the family life- cycle. Adventuresomeness III. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective adventuresomeness. Table 35 shows that the first group on in-home shoppers in the sample tend to be more adventuresome than all other in-home shoppers. Therefore, the above hypothesis does not appear to be validated by the data obtained from the sample, i.e., in-home shoppers who purchase from large catalogue stores, club-of-the-month organizations, and specialty retailers can probably be differentiated from all other in-home shoppers, when classified by their adventuresomeness. Cosmopolitanism IV. Ceteris paribus, among all in-home shoppers, there will be no difference, when classified by their respective cosmopolitanism. 152 The above hypothesis appears to be validated, as a result of the analysis of data on in-home shoppers. In fact, no differentiation was found in the sample between the first group on in—home shoppers and all other in—home shoppers, when classified by their cosmopolitanism. Trust in People V- Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective trust in people. Table 36 shows that the above hypothesis appears to be validated. The analysis of data on in-home shoppers, seem to indicate that no significant difference was found between the first group of in-home shoppers and all other in—home shoppers. Conservatism VI. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective conservatism. The results of comparison between in-home shoppers who purchase goods from large catalogue stores, club-of-the- month organizations and from specialty stores, and all other in-home shOppers seem to indicate that no difference could be found between them, as to their conservative-liberal attitudes. The above hypothesis, therefore, appears to be validated by the analysis of data from the sample. 153 Attitudes Toward Credit VII. Ceteris paribus, among all in-home shoppers, there will be no significant difference, when classified as to their respective attitude toward credit. The results of analysis showed that this hypothesis should probably be accepted. In fact, a comparison between the in-home shoppers classified in the first group, and all other in-home shoppers in the sample, indicated that no difference was found between them in terms of their respective attitude toward credit. Tendency for Impulse Buying VIII. Ceteris paribus, among all in-home shoppers there will be no significant difference when classified by their respective tendency for impulse buying. This hypothesis appears to be validated, as is shown in Table 35. A comparison of in-home shOppers who purchase goods from large catalogue stores, club-of-the-month orga- nizations, and specialty retailers and all other in-home shOppers in the sample showed that no difference could be found between the two groups as to their respective attitude toward impulse buying. 154 Second Group This group consists of in-home shoppers who purchase from novelty in-home shopping outlets, and who do not shop from specialty organizations. Social Class I. Ceteris paribus, among all in-home shoppers there will be no significant difference when classified by their respective social class. Table 35 shows that the above hypothesis seems to be validated by the results of the research. When the second group of in-home shOppers was compared to all other in-home shoppers in the sample, it was found that no significant differences existed between them, when classified by their respective social class. Occupation I. A. Ceteris paribus, among all in-home shoppers, there will be no significant difference, when classified by their respective occupation. As for the above hypothesis, the analysis of data shows that it should probably be accepted, with respect to the second group of in-home shoppers. In fact, no differ- ence was found between the in-home shoppers classified in the second group and all other in—home shoppers in the sample, when compared as to their respective occupation. 155 Income Level I. B. Ceteris paribus, among all in-home shoppers there will be no significant difference when classified as to their respective income level. The results of Table 35 shows that the above hypothesis appears to be validated. No difference was found between in-home shoppers of the second group, and all other in-home shoppers, when compared as to their respective income level. Educational Level I. C. Ceteris paribus, among all in-home shoppers, there will be no difference when classified by their respective educational level. A comparison between in-home shoppers classified as belonging to the second group and all other in-home shoppers in the sample, indicates that no differentiation was likely to be made by their respective educational level. The above hypothesis, therefore, seems to be validated. Family Life-Cycle II. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective stage in the family life-cycle. A comparison made between all in-home shoppers in the sample, and those in-home shoppers classified in the second group, indicates that the latter tend to be in the l“ 156 early stages of their family life—cycle, while the same is not true for all other in-home shoppers. The above hypothesis, therefore, should probably be rejected, and the alternative hypothesis accepted, i.e., it is likely that in-home shoppers can be differentiated from peOple who purchase goods at home from novelty retailers and who do not purchase from specialty outlets, as to their respective stage in the family life cycle. Adventuresomeness III. Ceteris paribus, among all in-home shoppers there will be no significant difference when classified by their respective adventuresomeness. Table 35 appears to indicate that no difference was found between in-home shoppers classified in the second group and all other in-home shoppers in the sample, when compared as to their respective adventuresomeness. The above hypothesis, therefore, seems to be validated with regard to the second group of in-home shoppers. Cosmopolitanism IV. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective cosmopolitanism. Results of comparisons made between in-home shoppers classified as belonging to the second group and all other in-home shOppers, seem to indicate that the former are of 157 less cosmopolitan orientation than all other in—home shoppers. This analysis points out that the above hypothesis appears not to have been validated, i.e., it seems possible to be able to differentiate among in-home shoppers, when comparing them as to their cosmopolitan attitudes. Trust in People V- Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective trust in people. As it can be seen in Table 35, in-home shoppers classified in the second group appear to have less trust in people than all other in-home shoppers. This tends to indi- cate that the above hypothesis should probably be rejected, i.e., the data analyzed seems to point out that it is pos- sible to differentiate among in-home shoppers when comparing their respective trust in people. Conservatism VI. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective conservatism. The above hypothesis appears to be validated by the results of the analysis. In fact, when all in-home shoppers in the sample were compared to in-home shoppers classified in the second group, no difference was found between them 158 as to their conservative—liberal attitudes. The above hypothesis, therefore, should probably be accepted. Attitude Toward Credit VII. Ceteris paribus, among all in—home shoppers, there will be no significant difference when classified by their respective attitude toward credit. Table 35 indicates that the above hypothesis is likely to be accepted. A comparison between the second group of in-home shoppers and all other in-home shOppers in the sample shows that no difference was found between them, as to their respective attitude toward credit. Tendency for Impulse Buying VIII. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective attitude toward impulse buying. This last hypothesis is also to be accepted. The analysis revealed that no difference was found between peOple who buy from novelty in-home outlets, and do not purchase from specialty retailers, and all other in-home shoppers in the sample, when compared as to their respective attitude toward impulse-buying. This hypothesis, therefore, appears to be validated. 159 Third Group This group consists of the in—home shoppers who purchase products from specialty and credit card in—home outlets. Social Class I. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified as to their respective social class. Table 35 indicates that in-home shoppers classified in the third group were likely to have higher social class than all other in-home shoppers in the sample. These results appear to invalidate the above hypothesis, so that the alternative hypothesis should probably be accepted, i.e., there seem to be no differences among in-home shop- pers, when classified by their respective social class. Occupation I. A. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective occupation. Individuals who shop at home from specialty and credit card type of outlets were found to tend to have higher status occupations than all other in-home shoppers in the sample. These findings appear to indicate that the above hypothesis was not validated by the analysis of the data, i.e., there appear to be significant differences among 160 in-home shOppers, when classified by their respective occupations. Income Level I. B. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective income level. Table 35 indicates that there appear to be signif- icant differences between in-home shoppers pertaining to the third group and all other in-home shoppers in the sample, when they are compared by their respective income level. In fact, individuals who shOp at home from specialty and credit card outlets are likely to have a higher income level than all other in-home shoppers. This indicates that the above hypothesis should probably be rejected, i.e., there seem to be differences among in-home shoppers when classi- fied by their respective income level. Educational Level I. C. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective educational level. A comparison between the third group of in-home shOppers and all other in-home shoppers in the sample indi- cated that no difference is likely to exist between them as to their educational level. The above hypothesis, therefore, appears to be validated. 161 Family Life-Cycle II. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified as to their stage in the family life-cycle. Table 35 indicates that the above hypothesis is probably validated by the findings. The comparison between the third group of in-home shoppers and all other in-home shoppers indicates that they could not be differentiated as to their stage in the family life-cycle. Adventuresomeness III. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective adventuresome attitudes. The in-home shOppers who purchase goods from Specialty and credit card outlets were compared to all other in-home shoppers in the sample, as to their adventuresome attitudes. It was found that the former tend to be less adventuresome than all other in-home shoppers. This appears to indicate that the above hypothesis should be rejected, i.e., that there seem to be differences among in-home shoppers when classified by their respective adventuresome attitudes. 162 Cosmopolitanism IV. Ceteris paribus, among all in—home shoppers, there will be no significant difference when classified by their respective cosmopolitanism. Table 35 indicates that no differentiation appears to be found between in-home shoppers of the third group and all other in-home shoppers, when compared as to their cos- mopolitan attitudes. The above hypothesis seemed validated by the analysis of the findings. Trust in People V. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified as to their respective trust in people. No difference appeared to exist between in-home shoppers of group three and all other in-home shoppers in the sample, as to their trust in people. The above hypoth- esis, therefore, seems to be validated by the research findings. Conservatism VI. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective conservatism. In-home shoppers who purchase goods from specialty outlets and credit card in—home retailers, were found not to be significantly different from all other in-home shoppers 163 in the sample as to their conservative-liberal attitudes. The above hypothesis is likely to be validated. Attitude Toward Credit VII. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective attitude toward credit. Table 35 indicates that the above hypothesis should probably be accepted. In fact, an analysis of the third group of in-home shoppers as compared to all other in-home shoppers in the sample shows that no significant differences appear to exist between them when classified by their respec- tive attitude toward credit. Tendency for Impulse Buying VIII. Ceteris paribus, among all in-home shoppers, there will be no significant difference when classified by their respective tendency for impulse buying. When in-home shoppers who purchase products from specialty stores and from credit card type of organizations were compared to all other in-home shoppers in the sample, no significant difference seemed to be found between them as to their tendency for impulse buying. The above hypoth- esis, therefore, is probably validated as to the research findings. 164 Analysis of the Research Findings in View ofithe Research Opjectives The objectives of the study are threefold. The first objective is to determine whether significant dif- ferences exist between in-home shoppers and individuals who do not shop at home. The second objective is to verify the existence of market segments within the in-home shOpping market. The third objective is to suggest methods for establishing segmentation criteria for firms in the in- home shOpping market in order to facilitate their selling operations. It is the Opinion of the researcher that the objectives of the study were accomplished in terms of the analysis of the findings. Accordingly, it was shown that: 1. There appear to be significant differences between individuals who shOp at home and those who do not. 2. Market segments can be delineated within the total in-home shopping market, with the aid of socio- economic and sociOpsychological variables. 3. The results of the research permit to suggest tentative methods of market planning for firms in the in-home shopping market, focused on the fact that segmentation is possible, and with the purpose of facilitating their marketing strategies. This aspect of the research will be eXplained in more 165 detail in Chapter VI, when the contributions of the study to marketing practice will be analyzed. Although not all hypotheses were validated by the results of data analysis, there is a clear indication that coupling socioeconomic characteristics with sociopsycholog- ical variables will accomplish a better delineation of the existing market segments. Summary Chapter V has presented an evaluation of the independent variables used in the research, as well as an evaluation of the research hypotheses. The last part of Chapter V provides a comparison of the research findings in view of the research objectives. The following chapter will provide a summary of the study, along with conclusions concerning its main contributions to marketing theory and marketing practice. Additional opportunity for future research will also be explored in the last chapter. CHAPTER VI SUMMARY AND CONCLUSIONS This chapter will attempt to summarize briefly the study, as well as indicate the main contributions of this research to the areas of marketing theory and marketing management. Areas suitable for further research in this field will also be explored, in order to indicate future Opportunities for more exact and detailed information on the subject. Summary of the Research The objectives of the study were to determine if individuals who shop at home, by mail or by telephone, could be differentiated from those who choose not to do so, and to determine if market segments exist within the in-home shop- ping market. The data was collected in the twin cities of Lansing and East Lansing, Michigan. A sample of one thou- sand subjects was sent the questionnaire and five hundred and nineteen answers were used in the analysis. Thus, within the limitation of sample size and location, several results were attained. In contrasting in-home shOppers with not in-home shoppers, it was found that: 166 167 1. Social Class: people who shop by mail and/or by telephone appeared to have higher social status, higher occupation and educational levels, and higher total family income, than those who do not shop at home. 2. Family Life-Cycle: in-home shoppers could not be differentiated from individuals who do not shOp at home by their stage in the family life-cycle. The marital status of in-home shoppers, the age of the head of the household, and the number of years married also did not significantly differentiate people who shop at home by mail or by telephone from those who do not. 3. Sociopsychological Characteristics: the results of the study indicated that in-home shoppers appeared to be significantly more cosmopolitan, less con- servative, and having a more favorable attitude toward credit than did individuals who did not shop at home. The variables, adventuresomeness, trust in people, and attitude toward impulse buying did not significantly differentiate the two respondents' groups. A second analysis of the data was made with the objective of discerning possible segments within the in-home shopping market. Three distinct submarkets were identified: 168 the first group was composed of those individuals who shop by mail and by telephone from large catalogue stores, club- of-the-month type of organizations, and from specialty in- home retail outlets. Those individuals were found to have high status occupation of the head of the household, high total family income, high educational level of the head of the household, high social class, and adventuresome attitudes. The second segment included respondent who stated that they utilized novely in-home retailers, and who did not shop at specialty retailers. The consumers in this group were found to be in the early stages of the family life cycle, to tend not to trust people, and not to be very cosmopolitan in their orientation. The third segment was composed of all those in-home shoppers who purchase their products from specialty and credit card type of organizations. The results indicated that the individuals in this group tended to have a high status occupation of the head of the household, high total family income, high educational level of the head of the household, and belonged to high social classes. They also seemed to display non-adventuresome attitudes. It is interesting to note that the main difference between the first and the third market segments is indicated by their adventuresome attitudes. Individuals of the first market segment appeared to be more adventuresome, while 169 those comprised in the third market segment appeared to have non-adventuresome attitudes. Conclusions from the Findingg An interesting conclusion that can be drawn from the findings is the fact that some of the variables that have proven to be of little use in differentiating in-home shoppers from those who do not shop at home, were useful in distinguishing market segments within the in-home shopping market. Such variables as family life-cycle, adventure- someness, and trust in people played important roles in delineating submarkets within the in—home shopping market. This indicates that it may be important to include for in- home shoppers market segmentation even those variables which cannot be used to characterize in-home shoppers as a homoge- nous group, or to differentiate them from not in-home ShOppers. The income level of consumers was shown to be a Strong influencer of the subjects' tendency to shop at home, as well as was their cosmOpOlitan attitude. Along with those variables, an important discriminator appeared to be the social class of the individuals. This appears to indi- Cate, within the limitations of the sample, that the more Educated and wealthier consumer is more prone to experiment With different purchasing practices than is the less edu- cated, less wealthy consumer. 170 At the same time, the respondents who tended not to accept local limitations and influences, seemed to be more prone to purchase from distant retailers, by mail or by telephone. The existence of less conservative attitudes in the in-home shoppers group may be related to their cosmo- politan orientation, to a certain extent. The tendency to accept different opinions and orientations, as well as different products, indicates absence of conservative, traditional attitudes, and at the same time appears to show the lack of geographical bounds. The reasons for the differences between the various market segments within the in-home shopping market may per- haps be found in the type of products offered by different outlets. Several variables such as store's image, the type of catalogues featured, and so many others could be respon- sible for differences in customers' characteristics. The objective pursued, however, was to prove that it is possible to determine the existence of segments within the market of in-home shOppers who purchase their products by mail and by telephone. This purpose was accomplished within the limitations of the sample used in the research. 171 ;/// Implications of the Research for Marketing Theory The fundamental contribution of the present study in the develOpment of consumer behavior theory is primarily descriptive. The consumer characteristics which resulted from the analysis of the data collected in the research yield significant contributions to the dimensions of dif- ferent segments within the in-home shopping market. The fact that the present research has descriptive qualities and appears to be an efficient discriminating tool among types of consumers, mainly those who purchase goods at 1 home by mail or by telephone, is a necessary condition to determine any predictive quality of it for the body of mar- keting theory. The ever increasing variety of consumer motivational forces affecting their behavior makes the job of defining and predicting behavior of market segments an ever more difficult one. Among the various qualities that are required to exist in marketing theory to justify its meaningfulness, are currency of data on which the predic— tions are based. In-home shopping is not a recent devel- Opment of marketing, however, urban in—home shopping is a very updated version of the phenomenon which started in the late 1800's. More recently, marketing theorists have been preoccupied with in-home shoppers, and attempts have been made to describe this particular type of customer. 172 Previous to this research, two studies in particular have attempted to establish differences between in-home shoppers and peOple who do not shop at home. The first study, by Peter L. Gillett, compared in-home shoppers with not in-home shOppers.1 Gillett examined demographic and socioeconomic variables, the influence of the fact that the consumer might be "locked in" the house, as a determinant of his in-home shopping activities, and, finally, whether or not in-home shoppers are more convenience oriented than those individuals who do not shop at home. Gillett found that the in-home shopper tended to have a better education, the head of the household had a higher status job, and the family's total income was greater than that of the not in- home shOpper. He also found that in-home shopping was not a consequence of the difficulty of getting out to the store. The third conclusion reached by Gillett was that the in—home shopper was more convenience oriented and flexible in his choice of alternatives than the not in—home shopper. Gillett concludes that: Urban in-home shoppers perhaps may be described as "modern" shoppers; they are seldom bound by shopping traditions and perceive less-than- average risk in buying by mail or phone. In- home shoppers are more affluent and better educated than other shoppers, but differ little on other major demographic characteristics.2 1Peter L. Gillett, "A Profile of Urban In-Home Shop- pers," Journal of Marketing, XXXIV (July, 1970), 40-45. 21bid., p. 45. 173 The second study, by Peters and Ford,1 also attempts to describe in-home shoppers. The major difference between these two studies is the fact that Gillett excluded from his research people who purchased from door-to-door salesmen, while Peters and Ford's study examines socioeconomic and personality characteristics of women who buy cosmetics from door-to-door sales representatives, and compares them with individuals who do not do so. The results of the second study furnished the following profile of the in-home customer: Relative to the woman who buys over half of her cosmetics in a retail store, the in-home buyer: 1. Has less access to a car for daytime shOpping; 2. Tends to be less educated; 3. Is likely to have more children living at home; 4. Is more likely to have a family income under $15,000 annually; and 5. The chances are greater that the head of the household will be a blue collar worker, clerical employee, or a salesman, rather than a professional.‘ Peters and Ford were unable to find significant differences between personality characteristics of in—home shoppers and individuals who do not shop at home. The present study should help clarify some of the differences that exist between Gillett and Peters and Ford 1William H. Peters and Neil M. Ford, "A Profile of Urban In-Home Shoppers: The Other Half," Journal of Marketing, XXXVI (January, 1972), 62-64. ZIbid., p. 64. 174 studies. The research was designed to determine if significant differences exist between individuals who shop at home and those people who do not, and determine if market segments can be located within the in-home shopping market. The findings of the study with respect to socio- economic variables seem to substantiate those of Gillett, and contract those of Peters and Ford. The active in-home shopper examined in the study, appears to tend to come from a higher social class, to have a higher family income level, and the head of the household to have a better educational level and occupation than those individuals who were clas- sified as inactive in-home shoppers in the study. The attitudinal test revealed that there was indi- cation that the active in—home shoppers were more liberal, had a more positive attitude toward credit, and were more cosmopolitan than the inactive in-home shoppers analyzed in the sample. Although these variables were not examined by either Gillett or Peters and Ford both previous studies try to distinguish attitudinal differences between the two groups.‘ Gillett had only limited success and Peters and Ford's study found no significant differences. The canonical analysis was able to distinguish market segments within the in-home shOpping market. This may imply that the findings of the Gillett and the Peters and Ford's study may actually not be in conflict with each other. While Gillett examined individuals who shopped at 175 home versus those peOple who choose not to do so, the Peters and Ford's study looked only at that segment of the market which purchased goods from door-to-door salesmen. Although the present study did not analyze door-to—door sales, it did conclude that distinct market segments do exist within the in-home shOpping market. In this context, the main theoretical contribution of the research will be found. Implications of the Research for ; Marketing Management i The research has shown, within the sample limita- tions, that the in-home shopping market can be segmented into groups with specific differentiating characteristics which include personality factors and emotional attitudes. The in-home shOpper has traditionally been classi- fied by mailing lists which are compiled either considering the individuals' socioeconomic and demographic variables, or the subjects‘ past purchasing behav1or with respect to mail and telephone purchases. The present study has chosen to examine the in-home shopping market from the perspective of combining socioeconomic characteristics with the individuals' emotional and psychological attitudes. If replications of the study, with different samples, indicate that persons with the same sociopsychological characteristics have sim- ilar behavior with respect to their purchase of products by mail or by telephone, then retailers should take into care- ful consideration this fact when making marketing decisions. 176 Although in-home retailing has been growing in importance for a long time, it is likely that promotional appeals which are aimed directly at the target markets would boost the growth Of the industry. It is necessary here to stress the fact that the research findings are limited to the sample considered, and to the time in which the study was realized. It is possible, therefore, that major variations exist as to the type of sociopsychological variables which may differentiate in- home shOppers from individuals who do not shop at home. Replications of the study would be necessary in order to verify the existence of such variations perhaps with the addition of other attitudinal variables to those considered in this study. The practitioner, therefore, should consider his own specific market, without generalizing to it the results Of this research. A better measure of segmentation would be to test his market for socioeconomic and sociopsychological differences since elements Of time and space have not been accounted for in this particular study. Evidence presented in Chapter II shows that the in- home shopping market presents currently a substantial and expanding market Opportunity. The urban shopper patronizes in-home sales for various reasons, as shown in Chapter V, among which are his own attitudes and motives. 177 This indicates that there are Opportunities for innovations within the in-home shopping market. Customizing products and services to consumers' demand is not new in marketing strategy. The size and heterogeneity of the in- home shOpping market suggests that such a policy may be successful in isolating segments which properly designed marketing strategies will help reach more effectively. The same evidence, in turn, may indicate with a fair degree Of accuracy how the in-home retailers should allocate their marketing efforts. The delineation of rela- tively homogeneous market segments with respect to relevant demographic, socioeconomic and sociopsychological variables derived by the research, enables tO assume that, when such characteristics are described with respect tO a specific in- home market, marketing efforts could be more efficiently allocated on the basis Of the relative Opportunities which exist in each submarket segment. The size Of each segment along with the historical value Of individuals' in-home purchases, might provide an initial indication of the apportionment of the marketing effort as a proportion of the market share represented by each submarket. Finally, one of the contributions Of this research to marketing management will be that of providing some criteria for the selection of marketing policies. If the marketing manager can clearly delineate the characteristics of each of his market segments, this will provide him with 178 an indication Of the appropriateness Of differing pricing, promotion, and distribution policies. The sociopsycholog- ical characteristics of consumers within each market segment could prove to be a valuable tOOl in providing an indication Of the appeals to be used in further analysis to obtain answers regarding the allocation of marketing expenditures. .. '51 In general, the dynamic marketer could continually use research tO investigate the determinants of demand for his own present and potential in-home market Offerings in order to maintain marketing effectiveness, improve his 1"“ competitive position, or even increase the size of his market. Suggested Areas for Further Research The study has not answered all the questions. It has primarily increased the number Of inquiries that pervade the marketing universe. The analysis Of in-home retailing has revealed that, once more, whenever human behavior is examined, no single solution or specific answer can be attained. The research has attempted to analyze a small number of variables which compose human behavior in order to determine which of these variables would play a relevant role in explaining the individual's purchasing behavior. Several topics for future research in this field can be noted. First, a similar study could be applied to 179 a larger sample in a different or several different geographical areas. It would be very relevant to verify whether the present findings would hold if the research were made on a national basis. Second, other studies could be made on specific in-home retailers. Observations about the behavior of a large sample Of catalogue store customers, for instance, might reveal the existence of influencing factors which were not included in this study hypotheses. Third, a longitudinal study of in-home shoppers could be done. Their behavior along the time would then be analyzed, revealing perhaps periodical changes in the purchase influence factors. Fourth, other emotional, personality, and motiva- tional attitudes could be tested. The list Of sociopsy— chological variables analyzed by the present study is far from exhaustive, and several more variables of this kind might increase the possibilities Of segmentation within the in-home shopping market. New scales Of attitudinal variables could also be tested, perhaps relating products to customers. The value of this study will be greater if it can be proved that sociopsychological variables can, in fact, lead to a more accurate segmentation of markets than that Obtained through the application of only socioeconomic and demographic variables. BIBLIOGRAPHY 9': .oh'( BIBLIOGRAPHY Alpert, Mark I., and Robert A. Peterson. "On the Interpretation of Canonical Analysis." Unpublished Working Paper 71-45. The University Of Texas at Austin, Austin, Texas, 1971. Anderson, Wilton T., Jr. "An Analysis of the Correlates of Convenience Oriented Consumer Behavior: With Special Emphasis on Selected Convenience Foods and Durable Goods." Unpublished Ph.D. dissertation, Michigan State University, East Lansing, Michigan, 1969. Barber, Bernard. Social Stratification. New York: Harcourt, Brace, and World, Inc., 1957. Bartolomew, Ralph. A Short Course in Direct Mail. New York: Direct Mail Division, Publishers Printing Co., 1928. Berkwitz, Leonard W. M. The Encyclopedia of Mail Order Business. New York: Leonard Berkwitz, 1900. Bigelow, Howard F. "Financing the Marriage." Family, Marriage and Parenthood. Edited by Howard Becker and Reuben Hill. Boston: D. C. Heath, 1948. Braid, Dorothy, "Individual Incomes and the Structure of Consumer Units." A.I.A. Papers Proceedings, May, 1958. Brandt, Steven C. "Dissecting the Segmentation Syndrome." Journal of Marketing, XXX (October, 1966), 22. Buckley, Homer J., G. D. Crain, and Maxwell Droke. Mail Order and Trade Paper Advertising. New York: International Textbooks Company, Bluebooks, 1937. Bull, Albert E. Mail Order and Instalment Trading. London: Sir Isaac Pitman and Sons, 1926. Burnett, Ed. "How to Select Lists that Reach People with Ability to Buy." The Reporter of Direct Mail Adver- tising, April, 1969, p. 34. 180 181 Clark, Lincoln, ed. Consumer Behavior II: The Life Cycle and Consumer Behavior. New York: New York University Press, 1955. Coleman, Richard P. "The Significance Of Social Stratifi- cation in Selling." Proceedings of the 43rd National Conference of The American Marketing Association. Edited by Martin L. Bell, December, 1960, pp. 171-184. Cooley, William W., and Paul L. Lahnes. Multivariate Procedures for the Behavioral Sciences. New York: John Wiley and Sons, Inc., 1962 Cunningham, William H., and W. J. E. Crissy. "Market Segmentation by Motivation and Attitudes." Journal of Marketing Research, IX (February, 1972), 100-102. Demby, Emanuel, and Louis Cohen. "Consumer Life Style More Vital to Marketing than Income Level." The Reporter of Direct Mail Advertising, June, 1969, p. 36. Dreesmann, A. C. R. "Patterns of Evolution in Retailing." Journal of Retailing, XLIV, No. 1 (Spring, 1968) 64-67. Doody, Alton F., and William R. Davidson. "Next Revolution in Retailing." Harvard Business Review, May-June, 1967, pp. 4-200 Duvall, Evelyn M. Family Development. New York: J. B. Lippincott Company, 1947. Duvall, Evelyn M., and Reuben Hill, co—chairmen. Report of the Committee on the Dynamics of Family Interaction. Washington, D.C.: National Conference on Family Life Cycle, February, 1948. Mimeographed. Egner, Frank, and L. Rohe Walter. Direct Mail Advertising and Selling. New York: Harper and Bros., 1940. Emmet, Boris, and John Jeuck. Catalogues and Counters. Chicago: The University Of Chicago Press, 1950. Engel, James F., David T. Kollat, and Roger D. Blackwell. Consumer Behavior. New York: Holt, Rinehart, and Winston, Inc., 1968. Engel, James F., David T. Kollat, and Roger D. Blackwell. "Personality Measures and Market Segmentation--Evidence Favors Interaction Views." Business Horizons, June, 1969, pp. 61-69. -T gr 182 Evans, Franklin 8. "Psychological and Objective Factors in the Prediction Of Brand Choice." Journal of Business, NO. 32, October, 1959, pp. 340-369. Fenvessy, Stanley. "The Dynamic Wholesale Mail Order Business: Evolution or Revolution?" The Reporter of Direct Mail AdvertiSLng, October, 1960, pp. 51-53. Ferber, Robert, Donald F. Blankertz, and Sydney Hollander, Jr. Marketing Research. New York: The Ronald Press Company, 1964. Foote, Nelson N., ed. Consumer Behavior: Models of Household Decision Making. New York: New York University Press, 1961. Foote, Nelson N., et al., eds. Housing Choices and Housing Constraints. New York: McGraw—Hill Company, 1960. Frank, Ronald E., William F. Massy, and Donald B. Morrison. "Bias in Multiple Discriminant Analysis." Journal of Marketing Research, August, 1965, pp. 250-258. Gillett, Peter L. "A Profile of Urban In-Home Shoppers." Journal of Marketing. XXXIV (July, 1970), 40-45. Glick, Paul C. American Families. New York: John Wiley and Sons, Inc., 1957. Glick, Paul C. "The Family Cycle." American Sociological Review, No. 14, April, 1947, pp. 164-174. Green, Roy M. "Think Creatively About the Lists You Use and the Markets You Sell." The Reporter of Direct Mail Advertising, June, 1967, pp. 16-17. Greenlaf, William. American Economic Growth Since l860. Columbia, S.C.: University Of South Carolina Press, 1968. Haer, John L. "An Empirical Study Of Social Class Awareness." Social Forces, NO. 36, December, 1957, pp. 117-121. Haley, Russel I. "Benefit Segmentation: A Decision- Oriented Research Tool." Journal of Marketing, XXXII Hawey, J. "What Makes a Better Seller." Motivation and Market Behavior. Edited by R. Ferber and H. G. Wales. Homewood, Ill.: Richard D. Irwin, Inc., 1958. 183 Higgins, Charles F. "Tfimz Booming In-Home Market." The Reporter of Direct Mail Advertising. Editor's Choice, Summer, 1967, pp. 48-49. Hodge, Robert W., Paul M. Siegel, and Peter H. Rossi. "Occupational Prestige in the United States." Class, Status and Power. Edited by Reinhard Bendix and Seymour Martin Lipset. 2nd ed. New York: The Free Press, 1966. Hollingshead, A. B. Elmtown's Youth. New York: John Wiley and Sons, Inc., 1949. Hollingshead, A. B., and F. C. Redlich. Social Class and Mental Illness. New York: John Wiley and Sons, Inc., 1958. Howard, James E. How to Use Mail Order for Profit. New York: Grosset and Dunlap, 1963. Howard, John A. Marketing Theory. Boston: Allyn and Bacon, Inc., 1965. Jain, C. L. "The Cost and Value Of Customers--How to Figure It for Yourself." The Reporter of Direct Mail Adver- tising, December, 1968, pp. 44-45. Jain, C. L., and A. Migliaro. "The Fundamentals of List Segmentation." The Reporter of Direct Mail Advertising, June, 1967, p. 29. Kahl, Joseph A. The American Class Structure. New York: Holt, Rinehart, and Winston, Inc., 1957. Kahl, Joseph A., and James A. David. "A Comparison Of Indexes of Socioeconomic Status." Social Forces, May, 1960. Kirk, Hazel. Economic Problems of the Family. New York: Harper and Row, Publishers, 1933. Kirk, Hazel. A Theory of Consumption. Boston: Houghton Mifflin Company, 1923. Kirkpatrick, E. L., Mary Cowles, and Roselyn Tough. "The Life Cycle of the Farm Family." Research Bulletin No. 121. Madison: Agricultural Experiment Station, University of Wisconsin, 1934. 184 Kotler, Philip. Marketing Management--Analysis, Planning, and Control. Englewood Cliffs, N.J.: Prentice-Hall, 1967. Lansing, John B., and Leslie Kish. "Family Life Cycle as an Independent Variable." American Sociological Review. No. 22, October, 1957, pp. 512-519. Lansing, John B., and James N. Morgon. "Consumer Finances Over the Life Cycle." Consumer Behavior II: The Life Cycle and Consumer Behavior. Edited by Lincoln H. Clark. New York: New York University Press, 1955. Lively, C. E. "The Growth Cycle of the Farm Family." Mimeographed Bulletin NO. 51. Wooster, 0.: Agricultural Experiment Station, 1932. Loomis, C. P. "The Growth of the Farm Family in Relation to Its Activities." Raleigh, N.C.: Agricultural Experiment Station, North Carolina State College, 1934. Loomis, Charles P. "The Study Of the Life Cycle Of Families." Rural Sociology, June, 1936, pp. 180-199. Martineau, Pierre. "SOCial Class and Spending Behavior." Journal of Marketing, XXIII (October, 1958), 123-125. May, Frederick E. "Buying Behavior: Home Research Find- ings." The Journal of Business, XXXVIII (October, 1965), 379-396. McCarthy, Jerome E. Basic Marketing. Rev. ed. Homewood, Ill.: Richard D. Irwin, Inc., 1964. McClosky, Herbert. "Conservatism and Personality." American Political Science Review, No. 52, 1958, pp. 27-45. McIntyre, Randall P. "Mail Marketing in Three Dimensions." The Reporter of Direct Mail Advertising, October, 1965, pp. 72-77. "More Firms Seen Selling Direct to the Consumer." Home Furnishing Daily, May 1, 1970, p. 7. Myers, James H., and Mark I. Alpert. "Detrimental Buying Attitudes: Meaning and Measurement." Journal of Marketing, XXXII (October, 1968), 13-20. 185 Nystrom, Paul H. The Economics of Retailing. 3rd. ed. New York: The Ronald Press Co., 1930. Osherow, Jerome B. "How to Clean a Mailing List." The Reporter of Direct Mail Advertising, October, 1960, pp. 39-40. Peters, William H., and Neil M. Ford. "A Profile of Urban In-Home Shoppers: The Other Half." Journal of Marketing, XXXVI (January, 1972), 62-64. Reid, Margaret. Consumers and the Markets. New York: Appleton-Century-Crofts, 1938. Reiss, Albert J., Jr., Otis Dudley Duncan, Paul K. Halt, and Cecil C. North. Occupations and Social Status. New York: The Free Press, 1961. Robinson, John P., and Phillip R. Shaver. Measures of Psychological Attitudes. Ann Arbor: Survey Research Center, Institute for Social Research, University of Michigan, 1961. Rodgers, Roy H. "Improvements in the Construction and Analysis Of Family Life Cycle Categories." Unpublished Ph.D. dissertation, Western Michigan University, Kalamazoo, Michigan, 1962. Rosenberg, M. Occupations and Values. Glencoe, Illinois: The Free Press, 1957. Sheth, J. W. "Multivariate Analysis in Marketing." Journal of Advertising Research, February, 1970, pp. 24-39. Shils, Edward A. "Class." Encyclopedia Britannica, Vol. V (1960), 766-768. Smith, Wendell R. "Product Differentiation and Market Segmentation as Alternative Marketing Strategies." Journal of Marketing, XXI (July, 1956), 3-8. Snyder, Ray. "The Mathematics Of Mail Order." The Reporter of Direct Mail Advertising, February, 1968, pp. 19-22. Sorokin, Pitirim A., Carl C. Zimmerman, and Charles J. Galpin. A Systematic Source Book in Rural Sociology. Minneapolis: University of Minnesota Press, 1931. Sparks, David L., and W. T. Tucker. "A Multivariate Analysis of Personality and Product Use." Journal of Marketing Research, February, 1971, p. 69. 186 Sparth, Frank W. "Classifying Direct Mail Lists." The Journal of Retailing, NO. 2, October, 1928, pp. 13-17. Sroge, Maxwell. "The Distribution Revolution . . . Anatomy of Direct Marketing." The Reporter of Direct Mail Advertising, July, 1970, pp. 18—20. Stevens, Milton S. "How Market Research Can Improve List Results." The Reporter of Direct Mail Advertising. Editor's Choice, Summer, 1964, pp. 27-29. Stone, Robert F. Successful Direct Mail Advertising and Selling. New York: Prentice-Hall, 1960. "Tele-Purchasing--Major Trend in Retailing?" Forbes, October 15, 1967, pp. 56—69. Terry, Pat. "More Customer 'Services' for Jordan March in 70's." Women’s Wear Daily, March 26, 1970, p. 29. "The Diners Club Story." The Reporter of Direct Mail Advertising. Editor's Choice, Summer, 1967, pp. 18-22. Tucker, W. T., and John J. Painter. "Personality and Product Use." Journal of Applied Psychology, No. 45, October, 1961, pp. 325-329. Tuey, Paul. Elements of Retail Salesmanship. New York: The McMillan Co., 1925. Udell, John G. "Can Attitude Measurement Predict Consumer Behavior?" Journal of Marketing, XXIX (October, 1965), 46-50. U.S. Bureau of the Census. Methodology and Scores on Socioeconomic Status, 1960, p. l. U.S. Bureau Of the Census. U.S. Census of Population, 1960. Subjects Report Socioeconomic Status. Final Report, PL(2)-5c. Washington, D.C.: Government Printing Office, 1967. U.S. Department of Commerce. "Mail Order Sales of Depart— ment Store Merchandise by Mail Order Companies or Mail Divisions." Survey of Current Business, February, 1962, pp. 5-11. U.S. Department of Commerce. "Mail Order Sales of Depart- ment Store Merchandise by Mail Order Companies or Mail Divisions." Survey of Current Business, LI, NO. 9 (September, 1970), 5-11. 187 Veldman, Donald J. Fortran Programming for the Behavioral Sciences. New York: Holt, Rinehart, and Winston, 1967. Wadsworth, Ralph K. Handbook of Mail Order Selling and Merchandising. Chicago: The Dartnell Corporation, 1918. Warner, W. Lloyd, Marcia Meeker, and Kenneth Eells. Social Class in America. New York: Harper Torch Books, Inc., 1960. Wundermann, Lester. "Mail Order, The Coming Revolution in Marketing." Advertising Age, July 24, 1967, p. 7. Wundermann, Lester. "Mail Order, The Coming Revolution in Marketing." The Reporter of Direct Mail Advertising. Editor's Choice, Summer, 1968, pp. 32-34. APPENDICES If") C"- I..' - l APPENDIX A LETTER OF INTRODUCTION MICHIGAN STATE UNIVERSITY EAST wsmc-mancm 48823 GRADUATE SCHOOL OF BUSINESS ADMINISTRATION DEPARTMENT OF MARKETING AND TRANSPORTATION ADMINISTRATION ° EPPLEY CENTER July 5, 1970 Mr. Bruce T. Allen 1876 Linden Street East Lansing, Mich. 48823 Dear Mr. Allen: I am a graduate student at Michigan State University working on a Ph.D. degree in Marketing. One of the requirements Of this degree is to complete a research project for my Ph.D. thesis. Enclosed in this envelope is a questionnaire which I am using as the basis for my research project. The project is designed to verify the existence Of submarket segments among people who purchase goods by mail or by telephone. To do this, I am asking a small sample of persons from a large cross-section of individuals to answer my questions. You are a member of my sample, and I need your help. Please assist me by completing the questionnaire enclosed with this letter. The questionnaire is relatively short and easy to answer. It is not identified in any way, so that it will not be possible to trace your answers back to you. I have supplied a stamped, addressed, return envelope to make it easier for you to mail the questionnaire to me when you have completed it. Your response is vital to my ability to complete a meaningful research project and I hope you'll take the time to complete this questionnaire and return it to me. I greatly appreciate your time and effort. Yours very truly, Isabella C. Mantovani Ph.D. Candidate 188 APPENDIX B FOLLOW-UP CARD ‘4‘ APPENDIX B FOLLOW-UP CARD A short time ago I mailed to you a questionnaire about in-home shopping. Since that time many of the questionnaires have been returned. If you have already completed yours I thank you for your cooperation. But if you have not yet responded, could you please do so at your earliest convenience? Your answers are vital to the results of my research. I realize that you are busy, but I still hope to have your cooperation. If you have any questions or comments, please feel free to call me. My telephone is: 355-5116. I greatly appreciate your help. Sincerely yours, Isabella C. Mantovani, Ph.D. Candidate 189 APPENDIX c QUEST IONNAI RE A SURVEY OF IN-HOME SHOPPERS DEPARTMENT OF MARKETING GRADUATE SCHOOL OF BUSINESS ADMINISTRATION MICHIGAN STATE UNIVERSITY EAST LANSING, MICHIGAN 191 INSTRUCTIONS FOR COMPLETING THE QUESTIONNAIRE All the inquiries recorded in this questionnaire may be answered by placing an "X" beside the response which you think affords the best reply to the question at hand (in certain exceptional instances you are asked to list items connected with the response). There are no inappropriate answers to any of the questions in this survey. For this reason, you should feel free to choose any response which you believe does the best job of depicting the answer most suitable to you. In order to provide the basis for a complete portrayal of the in-home shOpper, I have found it necessary to ask you for background informa- tion that you may View as being personal. Rest assured that your answers will be regarded as confidential. Furthermore, there will be no way in which your individual responses can be identified, or traced back to you. I appreciate your taking time to complete this questionnaire for me. 192 SECTION I Please answer the following questions by checking the statements that describe more appropriately your behavior and/or your Opinion. How many times do you go to stores to shop? (Don't consider grocery shopping.) ( ) once every day ( ) one to three times per month ( ) three times per week ( ) less than one time per month ( ) once or twice per week When you go shopping, who accompanies you? ( ) I usually prefer to go alone ( ) I usually go with my family ( ) I usually go with a friend or a relative Did you ever puchase goods by mail or by telephone? ( ) yes ( ) no Have you or any members of your family ever requested or ordered any of the following items by mail? ( ) membership in book clubs ( ) travel information ( ) membership in record clubs ( ) merchandise catalogs ( ) magazine subscriptions ( ) gift fruit/cheese or special ( ) household kitchen food products appliances ( ) clothing ( ) trees, plants, shrubs, ( ) film processing bulbs, flower/vegetable seeds Which of this year's general merchandise catalogs do you have in your home? ( ) Sears ( ) Montgomery Ward ( ) Spiegel ( ) J. C. Penney ( ) Alden's ( ) others, specify: , , Do you have any current gift or specialty catalogs in your home such as Spencer's, Sunset House, etc.? ( ) yes ( ) no If yes, please write the name of the catalogs: I I During the last two years, do you recall having purchased goods by mail or phone? ( ) yes, I have bought goods by mail or telephone regularly during ,the last two years. ( ) yes, I have bought goods by mail or by telephone occasionally during the last two years. ( ) yes, I have bought goods very rarely by mail and/or telephone during the last two years. 8. 10. 11. 193 ( ) no, I have not bought any goods by mail or telephone during the last two years. DO you purchase goods by mail or by telephone from large catalog or department stores like Sears and Roebuck, Montgomery Ward, J. C. Penney, Knapp's and others? ( ) yes, I buy regularly by mail and/or by telephone from those stores. ( ) yes, I buy occasionally by mail and/or by telephone from those stores. ( ) yes, I buy very rarely by mail and/or telephone from those stores. ( ) no, I have not bought any goods by mail or telephone during the last two years from those stores. During the last two years how many times have you bought products by mail and/or by telephone from large catalog or department stores? ( ) more than twice a month ( once every three-four months ( ) once or twice a month ( once every five-seven months ( ) once every two months ( less than 2-3 times/year ( no, I haven't bought at all VVVV Excluding large catalog and department stores, have you, during the last two years, purchased goods by mail or by telephone from any other organization? ( ) yes, I purchased goods from novelty catalog stores such as Spencer Gifts, Walter Drake, etc. ( ) yes, I purchased goods from "Club-of-the-month" type Of organizations (Doubleday Book Club, Columbia Record Club, Book-Of-the-month, etc.) ( ) yes, I purchased goods from specialty Mail Order Houses (Creative Playthings, Fredericks of Hollywood, etc.) ( ) yes, I purchased goods from credit card organizations (Oil Companies, Bank Credit cards, other credit cards) ( ) yes, I purchased goods from other types Of Direct Mail offers (magazine subscriptions, insurance, etc.) ( ) yes, I purchased goods which were advertised in magazines and newspapers, by using coupons or by telephone. ( ) yes, I purchased goods by mail or by telephone from other types of Operations. Explain ( ) no, I did not. If yes: During the last two Years how many times have you bought products by mail and/or by telephone from organizations other than large department and catalog stores? ( ) more than twice/month ( ) once every three-four months ( ) once or twice/month ( ) once every five-seven months ( ) once every two months ( ) less than 2-3 times/year 12. 13. 14. 15. 16. 17. 18. 194 Please indicate the name Of the organization(s) from which you purchase What is your marital status? ( ) single ( ) divorced ( ) separated ( ) married ( ) widowed If married, how long have you been married? (Please indicate the total number of years you have been married.) years. 3 the homemaker employed? ) full time (at least 35 hours/week) ) part time (less than 35 hours/week) ) I ( ( ( not employed The head Of the household is: ( ) 18-20 years of age ( ) 45-49 years of age ( ) 20-24 years of age ( ) 50-54 years of age ( ) 25-29 years of age ( ) 55-59 years of age ( ) 30-34 years of age ( ) 60-69 years of age ( ) 35-39 years of age ( ) 70 years of age or older ( ) 40-44 years Of age DO you have any children living at home? (Include only children presently living at home.) ( ) yes ( ) no If yes, How many 3 years Old or younger 0 l 2 3 4 5 How many 4-5 years of age 0 l 2 3 4 5 How many 6-12 years Of age 0 l 2 3 4 5 How many 13-15 years of age 0 l 2 3 4 5 How many 16-19 years of age 0 1 2 3 4 5 How many over 20 years of age 0 l 2 3 4 5 Do you have a TV set? ( ) yes, one ( ) yes, 2 or more ( ) no, I don't have a TV set Do you or anyone in your family have any hotel, restaurant, gasoline or other credit cards? (Indicate only the most used ones.) ( ) yes ( ) no If yes, please name the credit cards: Yourself: Other members of the family: 19. 20. 21. 22. 23. 24. 25. 26. 27. 195 DO you or anyone else in your family have charge accounts in department or apparel stores? ( ) yes ( ) no If yes, please name the stores: (indicate only the most frequently used.) Yourself: Other family members: DO you or other family members subscribe to any magazine? ( ) yes ( ) no If yes, please indicate only the name Of the magazines which you have subscribed most Often. Which, if any, of the following appliances are used in your household? ( ) automatic washing machine ( ) refrigerator ( ) automatic clothes dryer ( ) window air conditioner ( ) automatic dishwasher ( ) central air conditioner ( ) home freezer unit ( ) none of these DO you have a telephone in the house? (Do not include extension phones.) ( ) yes, one ( ) yes, more than one ( ) no, I don't have telephone How many automobiles are owned or used regularly by members of your household? (Count company cars kept at home.) ( ) none ( ) 2 automobiles ( ) 1 automobile ( ) 3 automobiles Who is the chief wage-earner in the family? ( ) husband ( ) wife ( ) others (please specify) What is his (her) occupation at the present time? (Please describe the type of job held.) What income-group best describes your total family income for 1960? ( ) under $5,000 ( ) $10,000 to $14,999 ( ) $5,000 to $7,999 ( ) $15,000 to $24,999 ( ) $8,000 to $9,999 ( ) $25,000 and over What is the highest level of education attained by the head of the household? Elementary High School College ()1&2 ()1 ()1 ()3&4 ()2 ()2 ()5&6 ()3 ()3 ( ) 7 ( ) 4 ( ) 4 ( ) 8 ( ) 5 or more 28. 29. 30. 31. 32. 33. 196 What has been the average total income per year of your household for the last three years? ( ) under $5,000 ( ) $10,000 to $14,999 ( ) $5,000 to $7,999 ( I $15,000 to $24,999 ( ) $8,000 to $9,999 ( ) $25,000 and over Some people say that most people can be trusted. Others say you can't be too careful in your dealings with people. How do you feel about it? ( ) Most people can be trusted. ( ) You can't be too careful. Would you say that most people are more inclined to help others, or more inclined to look out for themselves? ( ) TO help others. ( ) To look out for themselves. If you don't watch yourself, people will take advantage of you. ( ) Agree ( ) Disagree ( ) Undecided NO one is going to care much what happens to you, when you get right down to it. ( ) Agree ( ) Disagree ( ) Undecided Human nature is fundamentally cooperative. ( ) Agree ( ) Disagree ( ) Undecided 197 SECTION II A. Please indicate, by checking the appropriate box, whether you strongly agree, agree, disagree, or strongly disagree with each of the following statements: Strongly Strongly Agree Agree Disagree Disagree 1. It's best to join clubs where there are people mostly like yourself. ( ) ( ) ( ) ( ) 2. When it comes to what I do in my spare time I don't pay much attention to what people might think. ( ) ( ) () ( ) 3. In order to better himself and his family, a man sometimes has to give up some of his friends. ( ) ( ) ( ) ( ) 4. If given a choice between an American item and a foreign one I would select the American product even if the foreign one was cheaper. ( ) ( ) ( ) ( ) 5. People ought to pay more attention to new ideas, even if they seem to go against the American way of life. ( ) ( ) ( ) ( ) 6. If a person gets tired Of people he's known for years, he should stop seeing them. ( ) ( ) ( ) ( ) 7. Charge accounts facilitate the shOpping activity and avoid the nuisance of continually writing checks or carrying large amounts of cash. ( ) ( ) ( ) () 8. Charge accounts are only a nuisance and stimulate unnecessary buying. ( ) ( ) ( ) ( ) 9. Without installment credit we wouldn't be able to purchase many necessary items. ( ) ( ) ( ) ( ) 10. People generally pay interest rates on installment credit. ( ) ( ) ( ) ( ) 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 198 Credit cards organizations charge abusive interest rates on the purchase Of goods they Offer in installments. Credit cards are very convenient and don't cost anymore than cash. Advertisements in magazines and newspapers remind people of things they may need to buy. When you go shopping, you always end up buying something you hadn't planned to. It's not the little things we buy every now and then for pleasure that hurt our budget. If you go shopping with friends you end up buying unnecessary things just to be sociable. Life would be very dull if we only bought necessary things. The worst way to go shopping is to buy only the items of which you made a list before leaving home. I like to buy new products advertised on TV. I love unusual food and I fre- quently try new recipes. It is preferrable to take some risk and be the first one to buy a new product. It is really better to wait for some reliable feedback before you enter any new venture. I approve of people who don't fear social criticisms for experimenting with new ideas We buy new products before most of our neighbors do. Strongly Agree Strongly Agree Disagree Disagree 199 £052. woummmflo mauuflq m. mauuflu mmwwm. L052 xmv> adamumcmu woummmflo m .Naamuocwo Num> mmwmmmfio mmumc momma mcflonHOw 0:» mo some nufl3 Lose >uw> wwummmflp no .mummmmmxmnm no .cw>onvmmm no .cflmumcem mxfla GOmwmm ummum m mEOOOQ Op cofluwnfim may m>m£ mmEHUOEOm H .wawm>8 on cm>m menu weeps ou mxHH u.:0p H mHflnz .czo m.mco mm mama one mum mwmeawn cam mmummu mmos3 moumfioommm paw mpcmwwm xOflm Ou we m>fia ou >m3 ume use can mcoH 0:» GH .pmumsuu on ago 033 manomxm no mumpmma coma wamu on we so mcflom me umn3 zocx coo 03 >m3 >HCO any muoo mo OHHOB panaceameoo menu CH .mmop on mmcflnu meow mcu CH m>wflamn 0:3 mamomm one waveansm xomuum Op we uHEEOo cmo common m mEHuO umHOB one .moum u.cmo umsn H :Oflmmsomflp pmumw: 6 CA as pesos pom H mono C an“ “Om poom m.umn3 Bocx u.cov umsn mamomm umoz .muoxsanu umwum waammu mo dampen: m push coon >Hnmnoum O>mn mwmzu UGAxCME mo xuoumfln on» OH .umemucoo nummcmn ma mmwcammmz C30 was mo >Hflwmaflum mx:e£u 0:3 COmuwm c .muommmou Odo wmonu mo mcoflcflmo mcu was: ou wocmco m mm: wco Hausa co mcflom m.umn3 usonm ucmEmpsn m>uommu Ou manmuflmmp smumo ma uH .momHm >HO>OH wuuoum m we ca 0>HH 03 pauo3 on» .xaamucwEmpcsm "mangwumum .omummmflp xaamumcmm .Oauufla m concomfip .mauuwa m mwumm .owuom >Hamuwcoo .EOOE >um> momma 50> Hmcumsz .xon museumoummm on» mcflxomno >3 .ODMOHOGw mmmmam .OH 200 SECTION III A. This question is concerned with the type of products that you and your family usually buy by mail or by telephone. Below is a list of the kinds of products which you can buy from various retailers who sell through mail or telephone. They are alphabetically coded. We would like you to choose from that list of products, three of them and indicate: l--which type you buy the most frequently 2-—which type you buy less frequently 3--which type you buy even less frequently THE QUESTIONS TO THIS SECTION ARE AT THE BOTTOM OF THE PAGE. Alphabetical Product Alphabetical Product Code Types Code Types A Auto Supplies 0 Musical Goods B Household Goods P Office Needs C Clothing Goods Q Paints, Supplies D Baby Goods R Sporting Goods E Appliances 8 Novelty Goods F Electrical goods T Toys G Farm Supplies U Food Products H Floor Coverings V Books, Magazines I Gifts W Records, LP's, Tapes J Health—Beauty Aids and Cartridges K Jewelry X Furniture L Lingerie Y Shoes M Luggage Z Others (specify) N Notions I. Indicate below the types of goods you buy from retailers by mail or by telephone, in order Of frequency of purchase: (use alphabetic code) 1. 2. 3. II. The goods you usually purchase from retailers by mail or by telephone: A. Are goods which you could easily find in most retail outlets located near your home. B. Are goods which only a few retailers in this town have in their stocks. C. Are goods which are carried only by a small number of very Specialized retailers or by other retailers who sell by mail or by telephone. 201 III. The major reason for which you buy from retailers who sell by mail or by telephone is: A. NUDE! The price Of the goods. The convenience of buying without having to leave the house. The credit and/or charge facilities. The type of products Offered. Others (specify) B. This question has the purpose of verifying how most people spend their leisure time. Below is a list of activities people perform during their leisure time. Alphabetical Code Description A Watch television, listen to records or radio. B Visit with friends or relatives. C Work at special hobbies, or around the house or garden, weather permitting. D Read books or magazines for pleasure. E Go on short pleasure drives, window shopping, or picnicking, weather permitting. P Go to meetings or other organizational activities. G Go bowling, play pool, golf or tennis, swim, ski, or participate in other sports, weather permitting. H GO out to eat at a restaurant or snack bar. I Play cards, chess, checkers, or similar games. J Spend time at a bar, drug store, or similar place. K GO to see football, baseball, or basketball games, or other sports events. L Go to movies or drive-in movies. M Go to dances or parties. N GO to concerts, plays, adult school, museums, lectures, operas, or similar activities. 0 GO hunting, fishing, or camping, weather permitting. P Travel by plane, train, boat, or automobile. R Relax with my family, play with my children. S GO shopping with friends. T Others (please specify). We are interested in knowing how do you employ your leisure time: On weekday evenings On Saturday evenings On Saturday afternoons During eventual vacations II. III. 202 PLEASE LIST IN THE QUESTIONS BELOW, IN ORDER OF YOUR PREFERENCE, THE ACTIVITIES YOU USUALLY PERFORM DURING THOSE SPECIFIC PERIODS OF TIMES. LIST WHERE: The activities at which the most time is spent in each of the four primary leisure time periods. The activities at which the second most time is Spent in each Of the four primary time periods. The activities at which the third most time is spent in each Of the four primary time periods. (Use the alphabetical code to identify the activities.) On weekday evenings, I: On Saturday evenings, I: I I II II III III On Saturday afternoons, 1: During vacations, I: I I II II III III