.r .kxésx: ‘ . . . . ‘ . . . .. ). .Ik'. F I .. ,Iw. rl L‘tnrllllflxgii L I B R A R Y Michigan State _ -‘ . University 1x; r—Hx' .va- This is to certify that the thesis entitled POTENTIAL FAVORABILITY TOWARDS TELECOMMUTING: A PURCHASE DECISlONMAKING EXAMPLE presented by MAUREEN ELENA BENMINSON has been accepted towards fulfillment of the requirements for __P__.___h D degree in ,Masslledm' 47447221644. Major professor Date JAIL—“L, 1928 0-7639 T-”~".".“ 1'. mucwuumwfiivwnmrtfiOva ’OfivLCMOo .- V\ g a f a, Q ®e V . ~~i ,7 © 1978 MAUREEN ELENA BEN I NSON ALL RIGHTS RESERVED POTENTIAL FAVORABILITY TOWARDS TELECOMMUTING: A PURCHASE DECISIONMAKING EXAMPLE BY Maureen E. Beninson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY College of Communication Arts and Sciences 1978 CO! (I: (D [—4 ABSTRACT POTENTIAL FAVORABILITY TOWARDS TELECOMMUTING: A PURCHASE DECISIONMAKING EXAMPLE BY Maureen E. Beninson Background This study examined some of the human factors which might affect the adoption of a single in-the—home telecommuting service, purchase decisionmaking (TPDS). Telecommuting technology allows physical travel, result- ing in face—to—face interaction, to be replaced by tech— nologically mediated interaction via two-way communication systems. 'Previous empirical research has focused upon the teleconferencing or computer conferencing modes of telecommuting, as they relate to business operations. The psychological or behavioral reactions of individual consumers to nonoccupational, in-the—home telecommuting services have not received much attention. Using a knowledge base drawn from such fields as urban planning, communications, marketing and sociology, this study sought to uncover some of these consumer reactions.,/A purchase decisionmaking service was selected for investigation because retailing and banking Maureen E. Beninson industry trends seem to support more impersonal transaction modes; and if adopted, TPDS could supply the high volume of use required for interactive cable television system growth. Method A telephone survey of 261 adults in the East Lansing, Michigan area was conducted. The questionnaire was designed to allow two forms of investigation. First, a group activity pattern model, developed by Chapin, was used in an attempt to predict "favorability" towards TPDS. This model called for the inclusion of attitudinal, demo— graphic and behavioral characteristics as independent variables. The dependent variable, favorability, was measured in two ways, potential frequency of use and per— ceived value (amount willing to pay for service on a monthly basis). Second, information was gathered con- cerning how people use the impersonal communication— oriented (mail and phone) shopping services currently available to them. Factor analysis and various data manipulations were used to construct the research variables. Pearson product-moment correlations and standard multiple regres— sion analysis tested the variable relationships suggested by Chapin's model. Current impersonal communication— oriented shopping trends were assessed through the nu. Maureen E. Beninson calculation of chi—squares accompanied by Cramer's V, Spearman's rank order and Pearson's product-moment cor- relation coefficients. Results About two out of every five people were found to be fairly regular users of phone and mail shopping ser— vices. No outstanding differences were found between heavy and light users or the use of phone versus mail order services. The hypothesis tests revealed that people who were potentially willing to use TPDS more often were those: having more positive attitudes towards technology (1) (2) primarily concerned with comparing product prices and attributes (3) who wish to avoid conversations with sales clerks (4) who frequently conduct impersonal shopping trans— actions (5) who frequently shOp at night (6) who frequently drive to stores (7) who are younger (8) who are more mobile While people who were willing to pay more for TPDS were those: (1) who wish to shop alone " 5 1 $4: a BI Maureen E. Beninson who wish to avoid contact with salesclerks and (2) frequently combine shopping with other activities requiring travel (3) who frequently conduct impersonal banking trans— actions (4) who are less socially active (5) who are younger, of an earlier lifecycle status and a higher educational status (6) who are more mobile Conclusions The communication—transportation trade-off potential of TPDS, as it relates to travel reduction and more flexible scheduling of daily activities, appears to be a major factor in determining favorability towards TPDS. Also important is the question of how TPDS use may affect the satisfaction of social motivation, par— ticularly the need for interpersonal communication. Given the nature of telecommuting, it seems likely that such factors would affect any in—the-home telecommuting service. However, future research using a larger, more generally representative sample of people familiar with interactive cable operations is needed to fully assess these factors. n 1 . ‘ nq. a.) sl~ u... . ". ‘N. .13 I“ ‘ka' ‘1‘». \ m," “¥I . n 2- . ‘- .1 ‘N‘ 04“ ‘ l p U‘. ACKNOWLEDGMENTS I would like to express my deepest gratitude to my dissertation advisor, Dr. Martin P. Block, for his patience and guidance throughout this project. I would also like to thank the other members of my committee: Dr. John D. Abel, Dr. Charles K. Atkin, Dr. Gilbert D. Harrell and Dr. Gordon B. Miracle, as well as Dr. Thomas F. Baldwin, for their contributions to the development of this study. I am grateful to the management of East Lansing State Bank, particularly Stuart Bartlett, and all of the participating advertising students. Without their cooperation this study might not have been con- ducted. I am also thankful for the support I received at home from Greg and the watchful eyes of Yanu. Last, bUt not least, I must extend a special thanks to Dr. Kenward L. Atkin, who first introduced me to the World of advertising and always expressed faith in my ability to accomplish whatever I set out to d0- ii TABLE OF CONTENTS Page LIST OF TABLES. . . . . . . . . . . . . vi LIST OF FIGURES . . . . . . . . . . . . viii Chapter I. INTRODUCTION . . . . . . . . . . . 1 Conceptual Factors . . . . . . . . 5 Man's Environmental Needs . . . . . 5 The Current Environment . . . . . . 10 A Transportation—Communication Trade-Off Example . . . . . . . . . . . 12 The Need for Telecommuting . . . . . l6 Situational Factors . . . . . . . . 18 The History of Broadband Communications. 18 On Demand Communication Requirements. . 20 Media/Shopping Experiments . . . . . 26 Retailing Conditions . . . . . . . 3O Nonstore retailing trends. . 30 Store operation trends. . . . . 31 Delivery services . . . . . . 32 Banking Conditions. . . . . . 33 Growth of mechanization . . . . 33 The Service System. . . . . . 35 II. LITERATURE REVIEW AND DEVELOPMENT OF THE 49 HYPOTHESES . . . . . . . . . . . - ' . . 50 Shopping Versus Buying . . . . . 53 Product Differences . . . . . . . . III. IV. Social Participation . . . . . . Teleconferencing Behavior . . . General Classification of Reactions towards TPDS . . Selection of Independent and Dependent Variables . . . . . . . The Research Hypotheses . . . . METHODOLOGY. . . . . . . . The Research Technique . . Selection of Potential Independent Variable Questions . . . . . . Question Design . . . . . . . Demographic Questions Pretesting the Independent Variable Questions . . . . . Pretest Analysis. . . . DeveIOpment of the Dependent Variable Measures . . . . . Final Data Collection. . . . . The Sample. . . . . . The Interviewing Process . . . . Data Analysis. . . . . . Variable Construction . . . . Variable Composition . . . . 0 Demographic Independent Variables Behavioral Independent Variables Attitudinal Independent Variables Intervariable Correlation. . . . Hypothesis Testing . . . . . . FINDINGS . . . . . . . . . . The Sample . . Descriptive Questions—~Shopping by Phone or Mail. . . . . . Identification of Phone and Mail Shoppers. . . . Sources of Information. . . Type of Products Purchased . . Where Products Were Purchased iv Page 54 55 6O 64 7O 82 82 83 83 86 88 89 109 112 113 116 119 \\- “4‘. H‘. \\. \v’. 1“ Page Future Plans To Use Phone or Mail Shopping Services . . 119 Discussion of the Descriptive Question Findings . . . . . . . . . . 119 Dependent Variable Measures . . . . . 123 Bivariate Hypothesis Tests . . . . . 126 Summary of Attitudinal Bivariate Level Hypothesis Testing. . 132 Summary of Test Results for Hypotheses 4- 9, Demographic Characteristics . . 135 Summary of Test Results for Hypotheses 10— —l3, Media Use and Social Par— ticipation . . 139 Summary of Test Results for Hypotheses l4 and 15, Physical Mobility . . 141 Summary of Test Results for Hypotheses 16— —19 . . . . . . . . . . 145 Multivariate Hypothesis Tests . . . . 148 An Overview of Chapin's Model . . . . 156 Summary of Findings. . . . . . . . 160 V. CONCLUSIONS AND RECOMMENDATIONS . . . . 167 Background. . . . . . . . . . 167 Conclusions . . . . 172 Limitations and Contributions . . . . 138 Suggestions for Future Research. . . APPENDICES APPENDIX A. SURVEY QUESTIONNAIRE . . . . . . . . 184 B. INTERITEM AND INTERVARIABLE CORRELATION 196 MATRICES . . . . . . . . C. MULTIPLE REGRESSION STATISTICS. . . . 201 SELECTED BIBLIOGRAPHY . . . . . . . . 210 LIST OF TABLES Table Page l. POWER USED DURING PHYSICAL TRAVEL RELATED TO SHOPPING . . . . . . . . l4 2. POWER USED TO TRANSMIT ONE TWO- -WAY CABLE TV CHANNEL . . . . . . . . . . 15 3. SUMMARY OF THE LITERATURE: ATTRIBUTES WHICH MAY AFFECT REACTIONS TOWARDS TPDS. . . . . 65 4. TIMETABLE OF THE METHODOLOGICAL ACTIVITIES . . 84 5. DISTRIBUTION OF SAMPLE PHONE CALLS . . . . . 92 6. FACTOR LOADINGS AND AMOUNT OF VARIANCE EXPLAINED. . . . . . . . . . . 99 7. DEMOGRAPHIC CHARACTERISTICS OF THE SAMPLE. . . 110 8. PEARSON CORRELATION: PHONE AND MAIL SHOPPERS . 112 9. ALL MENTIONED SOURCES OF INFORMATION FOR PRO- DUCTS PURCHASED BY PHONE OR MAIL . . . . . ll3 10. CHI-SQUARE: FIRST MENTIONED INFORMATION SOURCE AND SHOPPING BY PHONE . . . 114 11. CHI- SQUARE. FIRST MENTIONED INFORMATION SOURCE AND SHOPPING BY MAIL . . . . . 115 12. PRODUCTS MOST RECENTLY PURCHASED BY PHONE OR MAIL . . . . . . . . 116 13. CHI-SQUARE: PRODUCT TYPE AND SHOPPING BY 117 PHONE . . . . . . . . . . . . 14. CHI—SQUARE: PRODUCT TYPE AND SHOPPING BY MAIL . 118 15. PEARSON CORRELATION: FUTURE PLANS TO SHOP BY 120 PHONE AND PAST USE. Vi l6. 17. 18. 19. 20. 21. 22. 23. PEARSON CORRELATION: FUTURE PLANS TO SHOP BY MAIL AND PAST USE . . . . . . . . . HOW OFTEN TPDS WOULD BE USED IF AVAILABLE FOR FREE . . . . . . . . . . . . MAXIMUM AMOUNT WILLING TO PAY FOR TPDS ON A MONTHLY BASIS . . . . . . . PRODUCTS MOST WANTED VIA TPDS. . . . . . PEARSON CORRELATION COEFFICIENTS FOR IMPER- SONAL SHOPPING FACTORS . . . . PEARSON CORRELATION COEFFICIENTS FOR INTER- PERSONAL SHOPPING FACTORS . MULTIVARIATE HYPOTHESIS TESTS: SIGNIFICANCE OF REGRESSION EQUATIONS AND AMOUNT OF VARIANCE EXPLAINED. . . . VARIANCE EXPLAINED BY SIGNIFICANT INDEPENDENT VARIABLES. . . . . . . . . Page 121 124 124 125 128 130 149 153 LIST OF FIGURES Figure Page 1. Chapin model . . . . . . . . . . . . 62 2. Data model based on Chapin's work . . . . . 71 3. Variance explained in TPDS usage by bivariate and multiple regression equations . . . . 158 4. Variance explained in TPDS value by bivariate and multiple regression equations . . . . 159 viii CHAPTER I INTRODUCTION Communications technology has long played a major role in the structure of our environment and daily activi— ties.1 Today, with the introduction of telecommuting tech— nology, whereby physical travel resulting in face—to-face interaction can be replaced by technologically mediated interaction via two—way communication systems, society is embarking upon an era within which radical lifestyle Changes are likely to occur. The identity of these changes has led to much speculation.2 Up to now, empir- ical research has primarily focused upon teleconferencing or computer conferencing modes of telecommuting as they relate to business operations or working conditions.3 The issues of whether these business applications are psycho— logically acceptable substitutes for face—to-face communi— cation, or if using telecommunications devices over long periods of time might also have behavioral or social consequences, have received very little attention.4 Questions in regard to how individual consumers might react to nonoccupational in-the-home services offered by e . I - l C .V «M v0 . flu Qt. ‘NM at. “It. tor is telecommuting technology, particularly interactive cable television, have been virtually ignored. This study attempts to examine some of the human factors which may affect the use of a single in—the-home service, purchase decisionmaking. The importance of considering the human factors of a new technology has been strongly demonstrated in recent years. Decisionmakers responsible for the instal- lation of the Lordstown, Ohio Vega Plant (which was sabo— taged by employees) and the Universal Product Code and automated checkout counters in stores (which have met with opposition from many quarters) learned the hard way that all Consumers are not ready and willing to accept all "technological improvements."5 /ghe problems experienced in the above situations are aptly analyzed in a telecommunications setting by Alex Reid in a 1971 report to the Sloan Commission. He Observed that while most telecommunications research is directed towards technological feasibility, this consid- eration is only a necessary, and not a sufficient, con— dition for the adoption of a new technology in a human communication system. Citing the human factor issues of need identification, need satisfaction and secondary effects or wide—range impact, Reid suggested that the shifting of efforts away from technological research towards studies of the human aspects of telecommunications is of the utmost importance.6 "(M -“"" The idea that a new communications technology i; not likely to equally satisfy the needs of all memberSLOf a society has received some support. Katzman, who has examined this issue from a cross-cultural point of view, has noted that: 1. People who first learn of or adopt an innovation have more education, higher status, more mass media exposure, more social participation, better interpersonal communication skills, and a more cosmopolite nature than those who find out or adopt later. 2. Higher informational (initial knowledge) status seems to be correlated with higher motivational levels to use new communication. 3. Higher socioeconomic status members are in a better position to acquire new communication technology. 4. The possession or lack of abilities required to merely use a new communication technique affects the value of its output to the receiver. The major social impact reported by Katzman is that new communication technologies have created infor— mation gaps between population subgroups. Viewing the dichotomy of high versus low economic and informational status, by the time lower group members have mastered a particular communication technique or technology, a newer technology has been adopted by the higher status group. In seeking a means to control the effect that an in—the— home telecommuting service could have on the current information gap between groups, human factors again prevail. The typical reaction of U.S. public policy- makers to such a problem has been to equate "equal access" to technology with the gap reducing power of equal use. Katzman suggests that equal access has failed to lessen the gap because motivational factors as well as other aspects of the psychological environment have not been considered.7 It would be difficult, however, to determine and effectively study all of the human factors associated with the broad range of potential telecommuting services, including education, medical care, business operations and Shopping in a single study. Thus, only one service area, purchase decisionmaking, has been selected. The act of purchase decisionmaking has been described in various models.8 A Telecommuting Purchase Decisionmaking Service (TPDS) will be defined as a service by which manufacturers, wholesalers or retailers may provide information to their customers, allowing them to do any of the following activ— ities: (1) To become aware of a product's existence (2) To know or understand what a product is and how it functions (3) To form a favorable or unfavorable attitude con— cerning a product (4) To decide whether or not they wish to purchase a product (5) To actually purchase a desired product h To PUr (6) To seek reinforcement for a purchase or non— 1: purchase deciSion The decision to study purchase decisionmaking as represen— tative of a telecommuting nonoccupational in-the-home service was made after close scrutiny of both the role of communications technology in society and the development of broadband communications, particularly the cable tele- vision industry in the United States. Support for this decision was also found within the current conditions of the retailing and banking industries. Conceptual Factors Man’s Environmental Needs Within primitive society, man's fraility compelled him to unite with others to satisfy his most basic needs of shelter, food and safety. The fact that exchanges of information were limited by man's ability to travel did not hinder man's life greatly since the state of strong interdependence made much travel prohibitive. Over time, developments in technology and science served to make ful— fillment of these basic needs less laborious, allowing man to become more independent. Man then came more strongly under the rule of transportation and communi— cation constraints because survival in conjunction with * . . . The value of this definition is that it allows people to examine product information without necessarily purchasing an item. 311 Eh: Specialization required exchanges of goods and services. As society moved away from a primarily agrarian status towards one of industrialization, cities, in order to facilitate these exchanges, became dense population centers. With the onset of further improvements in the means of transportation and communication, cities gradu— ally spread out, taking on the shape of a central business, finance and heavy industry core surrounded by layers of older manufacturing and housing, newer manufacturing and housing and finally reaching out into the remaining rural areas, bedroom suburbia.9 Today still greater improvements in communication and transportation technology which permit much flexi— bility in the levels of good and service exchanges are affecting the shape of our urban environment. Given the relatively high degree of personal control over our own living conditions which is now available, city planners and others are asking in just what direction the urban future should be shaped. Seeking to avoid the degradation of the mass, energy, space and information in the urban environment to an ultimate state of inert uniformity or entropy, several positions have been taken in identifying the most essential factor of urban life. Meier believes that the nature of city life is dependent upon the flow of information. He "sees the City essentially as the originator, conserver and \E- N t. Nu.“ AU transmitter of information." Placing an increasing rate of communication as the prerequisite of socioeconomic growth, Meier views the speed of information handling as the key to maintaining a city as a viable open system.10 Dakin, however, comments . . . what we have is an accelerating condition which feeds upon itself. The better the facilities and the better the standard of living, the more we shuttle bodies and goods about. We do not substitute one for one; we take the benefits of the information system, we pass around more information, and we also move ourselves more. Dakin, examining both information exchange and the move- ments of bodies and goods, has concluded that information exchange is probably less crucial than the latter.12 Entering this controversy from the fields of bio— logy and physical ecology, Odum and Odum conceptualize the City as an open steady state system through which energy flows. "The urban system is characterized by a faster flow——i.e., it has a more rapid metabolic rate than does the nonurban system or the rural region."l3 From their point of View, then, increasing the rate of metabolism or energy flow is essential. In applying these alternative solutions to life as we know it today, one quickly becomes caught in a web of interrelations between the entities of physical move— ment or transportation, communication and energy. The quest for a more efficient means of tranSpor_ tation for cities within the twentieth century has centered mainly around the automobile and various forms of mass transit. Getting started with the Highway Act of 1916, conceptual developments in the field of transpor— tation planning were relatively slow up until the mid 1950s when things began to accelerate quite rapidly. According to Webber, the major event occurring in the 19505 which stimulated growth in the field was the "rediscovery that land—use patterns and travel patterns are intimately related, a fact that was discussed in the early 19205, but then largely ignored.“14 The consensus of Mitchell and Rapkin, Carroll, Voorhees and Rannels was that . . traffic is a function of land use—~that traffic can be better understood and projected as a deriva— tive of activity-location patterns. Underlying this proposition is the more general idea that urbanites, being specialists, are dependent upon other urbanites with whom they communicate. Taking a slightly different tack than previous discussions, one mode of communication was then defined as the travel of individuals, another as goods shipment, expressed pri— marily as through truck, rail or air, and a third mode as message sending, through telephone, the mass media and other message handling channels. In this context, the city was redefined as a “vast communications system in which thousands of interdependent establishments, whether business firms or households or governmental agencies, interact with other establishments.“l6 In order to better understand life within this city, researchers began to collect data concerning temporal (duration, frequency, and timing of an activity) and Spatial (location prior to, during and following an activity) variables. Recognition of the value of these variable forms grew within the transportation field to the point where Kain acknowledged, "were it not for the fact that urban transportation demands are very unevenly spread over space and time, there would be no urban trans- portation problem."l7 Unfortunately, Kain's Statement no longer holds true. Added to the urban transportation problem, today, are potential energy limitations. In support of the con— clusion of the Ford Foundation's Energy Policy Project18 and projections which have been made by the Institute for Energy Analysis,19 a National Academy of Sciences Com— mittee interim report of a study commissioned by the Federal Energy Research and Development Administration indicates that in order to avoid adverse effects on the economy or American lifestyles in the future, a low rate of energy growth must occur.20 Rather than having to revert to a more primitive style of living, as many have feared, this report suggests that the key to achieving a low energy growth rate is to increase the level of efficient energy use and to change the mix of goods and services available towards those that require less energy. 10 The committee believes that the appropriate use of "tech- nological leeway" will allow for the existence of more energy efficient forms of production and transportation.21 Placing these implications for physical travel modes back into the realm of communication, the need for more energy efficient means of communication are also desirable, and energy use variables must be added to the transportation planning research scheme. The most essential factor of urban life at this point, then, seems to focus upon the speed and/or multi— tude of energy efficient modes of communication where transportation is considered one of these modes. Tele— commuting and TPDS, in particular, may allow society to communicate on a more energy efficient basis. The Current Environment Communication media consumption, if reflected in the number of available newspapers, periodicals and broadcast media stations, has generally increased since 1950. The total number of newspapers available has declined from 12,115 in 1950 to 11,292 in 1976, but circulation levels have increased.22 Daily papers which account for about 16 percent of all papers published, for instance, went from a total net paid circulation of 53,829,000 in 1950 up to 60,655,000 in 1976.23 The total number of periodicals has grown from 6,960 in 1950 24 to 9,872 in 1976. In the commercial broadcast area, “A 11 radio jumped from a total of 2,229 stations in 1950 to 2,573 FM and 4,388 AM stations in 1975;25 while the number of TV statiOns grew starting with 107 in 1950 to 719 in 1975.26 Cable television networks, within this same time period, expanded from 70 systems serving 14,000 subscribers to 3,366 systems with 9,800,000 sub— scribers.27 There are also several other modes of message delivery which are of interest. There were 15 million teletypewriter messages sent through the domestic tele- graph system in 1950. By 1974 the number of messages rose to 52 million.28 Penetration of the telephone into U.S. households grew from 62 percent to 95 percent between 1950 and 1975 with the average daily number Of all con— versations increasing from 176 million to 633 million.29 The number of domestic mail pieces being processed during this period also rose.30 Unfortunately, it is difficult to estimate how much energy is being used in the operation or consumption of all of these communication modes. In regard to both freight and passenger road traffic, the most recent volume of statistical abstracts reports that the average annual percentage of change in the number of vehicle miles travelled in the United States increased from a level of 3.9 percent with a total of 302.2 billion miles in 1950 to 6.9 percent and 1,268.3 billion miles in 1972. Then in 1973 while total 12 mileage continued to rise, the percentage of change began to drop and in 1974 a -l.5 percent change was experienced while the total number of miles travelled was lowered to 1,289.6.31 During this same time period, the volume use of railroads dropped but air travel increased. These travel patterns naturally affected the amounts of motor fuel, diesel fuel, jet fuel and elec— tricity used by these vehicles. However, when compared to other sources of fuel consumption such as industrial, commercial and household operations, transportation between 1970 and 1975 has maintained a steady draw of 25 percent of the total fuel available for use in the 32 It is this 25 percent factor which may United States. have to be reduced further by more energy efficient means of transportation in the future. TPDS and other tele- commuting services, by offering a more energy efficient means of communication, may be the key to achieving such a reduction. A Transportation—Communication Trade—Off Example In order to provide some basis for comparing transportation and communication energy use, an attempt has been made to estimate the difference in the kilowatts of power used by 20,000 people to travel to and from stores versus the use of an interactive shopping service carried on a 20,000 subscriber cable television system. 13 This estimate does not account for the amount of power being consumed by the stores people might actually shop in, or those stores' warehousing facilities, as opposed to the warehousing systems which could be operated in conjunction with an interactive shopping service. Assum- ing that the amount of energy or power consumed by a warehouse, plus any equipment required by the interactive system, is less than or equal to the amount consumed by a store, the absence of store and warehousing consider- ations biases the estimate in favor of physical travel. The calculations indicate that over a one—year period, 20,000 people would use an average of 869,000 kilowatts of power to physically travel to and from Stores (see Table 1). Running on a 24-hour, daily basis, a two-way cable television channel serving 20,000 people would, during a one-year period, use 46,000 kilowatts of power (see Table 2). Substitution of a two-way cable shopping service would result in a savings of 823,000 kilowatts annually. Including the power consumption of a delivery service which would be required by a cableshopping service requires further speculation. If a delivery truck were to average 5.3 miles of travel per delivery,33 .09 horse~ power or .07 kilowatts would be consumed per average delivery. In order to close the gap between the power consumption levels of shopping via physical travel and 14 TABLE 1 POWER USED DURING PHYSICAL TRAVEL RELATED TO SHOPPING Given: (1) The average person spends 2 hrs/week in shopping travela (2) An average driving speed of 35 mphb (3) One kilowatt (K) = 1.3410 horsepower (P)C (4) Total P'of Vehicles Used = 14,306 milliond _ .016 Total Mileage Travelled 887,800 million P/mile The shopping related travel of one person: 2 hrs x 35 mph = 70 miles per week 70 miles per week x 52 weeks = 3,640 miles per year 3,640 miles per year x .016 P/mile = 58.24 P per year 58.24 P _ m -— 43.43 K per year The total amount of power used by 20,000 people for shop— ping related travel during one year: 43.43 K per year x 20,000 people = 868,605.5 K per year aF. Stuart Chapin, Jr., Human Activity Patterns in the City: Things People Do in Time and in Space (New York: John Wiley & Sons, Inc., 1974), p. 99. (Figure based upon a 1968 sample.) bAn assumption made by the researcher considering normal high and low speeds in city and country driving. CCharles D. Hodgman, Handbook of Chemistry and Physics (Cleveland: Chemical Rubber PublishIng Co., 1960), p. 3169. dBureau of the Census, U.S. Department of Commerce, Statistical Abstract of the United States 1976, Tables No. 902: Total Horsepower of All Prime Movers and No. 986: Motor—Vehicle TraVel, By Type of Vehicle (1965 Figures), Pp. 547, 592. 15 TABLE 2 POWER USED TO TRANSMIT ONE TWO-WAY CABLE TV CHANNEL Given: (1) A one-way cable system having 20,000 subscribers delivers 15 channels over 450 miles of cable with 38,000 Kilowatt hours permonth.a (2) Upstream transmission requires 33 percent more power than downstream. The amount of kilowatts used per hour by the entire system: 38,000 K per month 720 hrs per month = 52'77 K per hour A single one—way channel uses: 52.77 K per hour 15 channels = 4 K per hour One hour of upstream transmission on 1 channel would require: 33 percent of 4 K = 1.3 K One hour of two—way transmission on 1 channel would require: 4 K + 1.3 K = 5 K One two-way channel running continuously for 1 year would use: 8760 hrs per year x 5.3 K = 46,428 K aInformation released to researcher by Local cable television system. bAn assumption made by the researcher given knowledge of two—way transmission technology. §\~ 16 an interactive cable TV service, a total of 11,757,142 deliveries would have to be made during that year. This averages out to 588 deliveries, more than 1 per day per person. Assuming that such a high level of deliveries is not likely to be required, the cable service would still consume less power. Whatever the true energy tradeoff level may be, it is relatively safe to assume that a single two—way cable television system, if equipped to handle mail, and news delivery along with a library access for periodicals and the presently more common ability to carry entertainment programs, could concurrently deliver these services to a large number of people more efficiently than the channels used today.34 The Need for Telecommuting The idea of substituting physical travel with some other communication mode is not a new one. Years ago, if people wanted to see something or talk to some— one, physical proximity to that person or object was required. Today via the mass media, the American public can see and hear things and people which they may never have in physical contact. Underlying these communication Style changes has been man's ability to substitute physi— cal travel with the use of machines.35 Yet relative to the increasing communication and transportation demands which have accompanied today's lifestyles, this reduction 17 in physical travel has not been that great. There are still many activities favored by people in which travel plays an integral part. The major difference between these activities and those which have been successfully replaced by mass media use is that they are receiver—initiated, sensory transactions involving goods, services and ideas. That is, people are choosing to physically participate in exchanges of goods, services and ideas at particular points in time. It is not difficult to determine why the impact of mass media and the telephone upon travel propensities has thus far been relatively low. The mass media, while expanding the horizons of their audiences, do so at their own convenience. The timing of radio or television programming, for example, is primarily con— trolled by the broadcaster, not the audience members. Interaction between audiences and the media, in the form of feedback, is limited to such things as letters, phone calls, and research determined show ratings. The tele— phone does allow the user to control the timing of activities but cannot necessarily provide all of the Stimuli needed. With the exception of the limited use Of Bell's Picturephone system, and some teleconferencing networks, one cannot receive or send audio—visual infor— mation via telephone.36 Ill 18 The next logical step in communications technology has been to attempt to facilitate audience control over both the timing and the content of the messages received. Pool, Murray and Dobb have labelled the concept of receiver-initiated electronically transmitted messages "on-demand" communication.37 It is through telecommuting technology that on—demand communication is likely to become available. Situational Factors The History of Broadband Communications In 1948, Community antenna television, often referred to as CATV or Cable TV, was developed to serve communities that were isolated from normal television service by geographical conditions and distance.38 Antennas, placed in high areas, were, and still are, used to pick up regular, over—the-air television signals. These signals are conveyed to a head end (the cable station) where they are fed onto the station's coaxial cable distribution system. This distribution system generally consists of trunk amplifiers transmitting the signal down trunk cables which branch out along streets to bridger amplifiers hooked up to a feeder cable system having feeder amplifiers. These feeder amplifiers deliver the signals to tap off units and then by service drops into subscribers' homes.39 The cable itself is normally strung upon utility poles or buried underground. \(‘1 Va] ~66 Prc tWO int. .0! 19 Cable systems were allowed to flourish without the threat of federal or state regulation until the mid 605. Then in 1966, in reaction to TV broadcasters' fears about CATV competition, the Federal Communication Commission (FCC) prohibited further importation of distant signals into the one hundred largest United States television markets. This action slowed down the CATV movement, but systems outside the major markets continued to grow espe- cially where, operating as middlemen between noncable television stations and consumers, they could offer better reception, and provide a broader range of channels than local TV reception. A third factor contributing to CATV growth was the offering of programming unavailable else- where such as automatic weather reports or local sports. The "freeze" ended in 1972 when the FCC issued its "Cable Television Report and Order." The rules presented in this report allow all CATV operations to bring in some distant signals while requiring systems in the major markets to provide at least twenty‘channels as well as various nonbroadcast services including channels for lease, individual channels for public access, educational Programming and municipal services, and a capacity for two—way services. This two—way requirement has been interpreted as a technical capability of sending nonvoice Signals upstream (from subscriber to headend) in addition to normal downstream operations.4O () 11) 20 Although, due to reasons to be disaussed later, not much progress has really been made in the establish— ment of operating two—way systems, CATV in general has spread from rural districts to all over the United States. There are now 3,500 operating cable systems, serving 7,800 communities. These systems reach about 10 million sub— scribers or 15 percent of the nation's TV households. There are 2,650 systems which have been approved but not 41 yet built. Optimistic reach projections for 1980 range from 42 percent of the United States TV households (Baer)42 to between 40 percent and 60 percent (Sloan Commission).43 On Demand Communication Requirements The question of when the general public will actually be able to utilize on-demand communication ser— vices has come under much speculation.44 The recent con- cern over energy conservation certainly strengthens the appeal of this potentially travel reducing, energy saving technology. Yet, as in the case of many new things, the realities of production, distribution and use costs must be dealt with. Estimates of the actual dollar cost of two-way CATV systems, in general, vary initially according to the type of cable equipment, switching system, computer facilities and in-home terminals being used relative to the progressive penetration of the total system into a geographical area. Added to this must be any software 21 costs which may be passed on to the consumer, as well as potential costs per unit of service used, these being derived from projections of the service unit duration and frequency.45 In more general terms, a major issue is what Baran has identified as another case of the clas— sical chicken and egg problem. "Without the availability of hardware there can be no market for services. And, without an economically viable market for two-way service offerings, low-cost availability of well—designed support hardware is unrealistic, for one without the other is worthless." To achieve maximum economies of scale, sys— tem compatibilities will be necessary on a nationwide basis. This necessitates uniform standards for equipment and interface systems.46 The government, while appearing to help this situ- ation in terms of setting performance and service level requirements, has in the long run been hindering two—way CATV growth. Accused of getting in the way of expanding CATV services have been government limitations on the entertainment portion of current CATV offerings which thereby reduce the potential revenue of CATV stations. Financial difficulties have also been enhanced by the recent recession, high interest rates and industry over— expansion. Particularly prohibitive to investment has been the lack of credibility now held by two—way CATV due mainly to early over-zealous speculation and over-selling 22 of the concept before the technology was available.47 Today, while much technology is available, there are still improvements needed, especially in regard to offering TPDS. One way of getting around many of these problems would be for the government to provide 100 percent pene— tration of homes in order to allow national polling, edu— cation or other public services. However, the more probable course in this country will be reliance on com- mercial processes.48 Upholding this view, Block, Wright and McVoy have put forth an evolutionary approach to the development of two-way cable technology communications consisting of a series of system generations. They believe that initial two—way CATV penetration can be achieved by offering rela— tively inexpensive systems which have simple monitoring capacities permitting per—program pay TV or low level security services (1st generation). Past experiences suggest that initial penetration of such a system is fairly high (40 to 80 percent), followed by a drop (25 to 40 percent) after the first few months of operation.49 However, as a system refines its programming, the revenue generated from the remaining subscribers grows and "lost" or new subscribers may be drawn into the system. Contributing to the pull power of the programming can be such interactive services as in—service training, 23 public forums, and continuously running catalogs of pro— ducts or information. Changes in the system, to provide the delivery of such services on a prearranged schedule and the active response situation of multiple choice questions (2nd generation), increase costs; but the shared access of these services (as opposed to on—demand communications) can keep these expenditures low enough that they be covered by the leasing of channel space to the service sources.50 As the number of services on a system grows, the problem of distinguishing between service signals is expected to arise. Further changes made in the system, which permit the identification of service signals through grouping and sequencing the transmission of data, also provide for the transmission of more complex data mes- sages (3rd generation). If certain interactive channels are given high speed transmission capabilities, a linkage similar to an on—line computer terminal can be established, thereby permitting services like electronic funds transfer. To achieve on—demand communication forms, addi- tional improvements in CATV systems will be required. Needed will be more complex information storage and trans— mission operations including frame grabbing (transmission Of individual still pictures) and hopefully digital—analog Signal transformations (providing more detailed message 24 display). Such changes might be expected to occur within or after the 4th generation.51 Assuming the inclusion of whatever equipment needed to obtain the greater storage and frame grabbing capacities, the basic system permitting on—demand communi- cations will be based on area multiplexing utilizing digitally controlled, code-operated switches (COS) which allow the separation and movement of downstream and up— stream signal frequencies. By placing a primary COS at each trunk and a secondary COS at each bridge amplifier of the traditional tree network, "polling" of service requests or responses by a minicomputer can occur.52 This will require a minicomputer having a large core memory and disk storage capacity. The size of a TPDS system will be dependent upon interconnections with other systems. Microwave relay systems, comprised of series of microwave repeaters, allow regional connections. For example, the Target Network Television System, originating in Kansas City, Missouri, currently provides three hundred cable systems in a twelve—state area with original programming via direct connection (trunk cables) and microwave.53 Communication satellites have the capacity to transmit signals within a larger geographical area, thus potentially extending CATV services even further.54 Looking back at the steps in the evolutionary pro- cess of two—way CATV development, it is clear that once 25 initial contact is achieved, it will be the demand level for specific services that will determine the speed of system growth. Having reached this level of consider- ation, the issue then becomes what kind of service will enough people be likely to perceive as sufficiently useful and economical to instigate a rapid rate of widespread adoption? A broad range of potential services to the home have been put forth by technicians and researchers.55 These can probably all be placed into one of the following categories: (1) education (2) community and medical services (3) protective or security services (4) general information (news, libraries, etc.) (5) shopping facilitation (6) business conducted from the home (7) entertainment (8) person-to-person communications The one service to the home which may meet a sufficiently universal need and which also has obvious receiver advantages linked to system improvements, there— by increasingly promoting system growth, comes out of the category of shopping facilitation——purchase decision- making.56 One advantage which this particular service has over most others is the public's relatively high 26 consumption rate of goods and services.57 If frequency of demand for the total purchase decisionmaking service can be expected to be fairly high, then the costs to the individual user will be lower. Once consumers are hooked up to the equipment necessary to adopt this service, many of the other services will be sure to follow, accelerating the rate of cable penetration. Collectively, a second advantage is represented by a number of trends present in the retailing and banking industries. These will be discussed later in this chapter. A third advantage is that there has been some indi- cation of consumer willingness to adopt such a service in the fact that three media—oriented shopping systems have been successfully tested or operated during the last four years . Media/Shopping Experiments Sears—Simpson and IBM of Canada ran a three-month test using product catalogs in combination with a computer— ized order processing system, that converted the unused buttons on customers‘ touch tone telephones into small computer terminals, enabling the placement of catalog orders. They reported that 453 out of 2,000 regular catalog customers placed approximately 1,130 successful orders.58 Call—A—Mart, a supermarket in Louisville, Kentucky, has been operating a system where customers verbally phone 27 in their catalog orders by using product identification numbers. The system's computer in addition to "recording the order, prepares an invoice, keeps a running tab and determines the optimum warehouse to home route for delivery truck drivers." A high initial interest was expressed by consumers in that 8,500 of them signed up to receive the $5 per week catalog during the first four months of service operation.59 In 1974, CableMart, Incorporated reported four and one—half years of success for their Cable Family Shopping Center. The Center, utilizing a combined mail and cable television delivery of product information and phone or mail product ordering, had up to that point broken even in its dealings with an eighteen-state cable network com— prised of 350,000 subscribers.60 Cable Catalog (Neiman—Marcus), GiftAmerica (Western Union) and Shop—at-Home (Cable Response Network, Incor— porated) have also used television product information delivery and either phone or mail product ordering systems in recent years. Consumer reactions to these programs cannot, however, be assessed because usage or sales figures have not been made public.61 Unfortunately, there have not yet been any tests which actually involved the use of two—way cable tele— Vision for both product information delivery and placement Of product orders. Between 1970 and 1975, several 28 two-way cable operations designed field studies to test shopping among other services: Tocom in Irving, Texas; American TV and Communications in Orlando, Florida; Theta- Com in El Segundo, California. However, all of these tests were cancelled or delayed due to unfavorable economic con- ditions.62 Since 1975, two—way cable technology has been suc— cessfully used to provide various services to people in Rockford, Illinois and Columbus, Ohio. The first appli— cation of the cable system in Rockford was to carry the ‘proceedings of city council meetings upstream to the headend, which then cablecast it live over the entire system. Researchers from Michigan State University, operating under a National Science Foundation Grant, ran an experimental training program for firemen in 1976—77. Firemen, individually or in groups, were able to view videotaped lessons on television sets located in the fire stations. During and at the end of each lesson, retention of information was tested by way of multiple choice questions. Firemen responded to the questions by pushing buttons on a modified channel converter console. The system's computer kept a record of the answers and pro— duced "test scores" at the end of each lesson. This project is significant to TPDS in that it successfully permitted Viewers to access and interact with the cable system via downstream signals.63 Columbus has two cable 29 systems which are using two-way cable technology. Coaxial Communications offers a service called Telecinema to approximately 24,000 subscribers. Telecinema is a form of pay TV. That is, given a broad range of movies to choose from, viewers, after an initial service fee, pay only for the movies they actually watch. This system is a technological improvement over many cable systems because of its ability to poll channel use on a minute— by—minute basis.64 Warner Cable Corporation, serving 24,300 residents of Columbus, tested a game show program which actually involved viewer participation in 1976. Viewers were asked to register their answers to questions being asked of the on—air participants by pressing buttons on their consoles. The Warner computer tallied the answers and flashed the response of the majority on the screen. Since the game show was well received, Warner has announced plans to expand its ten—channel capacity to thirty channels in order to provide such services as: 1. On—the—air sales for retailers 2. Rosie's Tag Sale, an electronic garage sale in which viewers will be able to bid on one another's unwanted goods 3. Educational courses for credit 4. Computer monitored games 5. A home fire and burglar alarm system 6. An all day locally produced version of the Today Show including audience participation in dis- cussions6 If Warner is successful in its expansion, it will be the first two-way cable system to offer any sort of shopping Service on a regular basis. 3O Retailing Conditions Nonstore retailing trends. Consumer use of non— store retailing operations, where physical travel to stores is replaced by mail or phone orders, appears to be rising. Mail order retailing, although started in the late 18005, received little attention until the early 19005 when Sears, Roebuck & Company and Montgomery Ward & Company produced extensive catalogs. Mail order use grew rapidly due to these and other companies' concentrated efforts within rural locations which could not financially support traditional retail outlets.66 Today traditional retailers have won over most of these markets, but at least 4 percent of total retail sales (approximately half of all nonstore retail sales) can be attributed to mail orders stemming from catalogs and media advertising.67 The catalog order business seems to be doing particularly well as evidenced by the increase in catalog houses from 2,000 to 5,000 between 1960 and 1970.68 It has been reported that the people who tend to be using mail order services are: . affluent more highly educated women with children (less mobile) suburbanites over city dwellers those who value shopping quickly 01.5me The volume of phone order retailing, which is also often linked to catalogs or media advertising, increased an average of 36 percent between 1950 and 1961 31 in eight major United States cities. Recently phone order retailing has received much retailer support due to the desire to maintain past business levels while cutting back store hours in reaction to the energy crisis.70 Another concern of retailers is the potential decrease in customer flow which can result from higher gasoline prices and changes in public travel activity.71 The primary users of phone order services appear to be similar to mail order users . 72 Store operation trends. The service characteris— tics of store operations have adjusted over the years in order to accommodate labor costs on the part of retailers; and the price consciousness, coupled with a desire to examine merchandise leisurely, making selections based upon personal judgment without the help of salespersons, of consumers.73 Thus, today while some "full service" stores, where customers receive highly personalized attention, still exist, "self—selection" designs, such as discount stores where individual salespeople are available only if needed, and "self—service“ stores, including supermarkets, warehouse retailers and catalog showrooms which basically provide a centralized customer-store employee interaction for checkout or order—taking and checkout, are more predominant.74 32 The warehouse and catalog showroom concepts are of special interest because these recent store designs reflect consumer acceptance of a longer hours available, but more impersonal, shopping experience.75 From a retailing point of View, such centralized operations alter merchandizing and inventory management needs and may be more readily adapted to full computerization. Approaching catalog mail orders as a specialized form of warehouse retailing, a German firm has reportedly pro— cessed 2 million items in eight hours within a single warehouse through the use of a computerized order taking-— mechanical and electronic product moving system. The only points at which humans become involved are in the final packing and delivery arrangement stages.76 Delivery services. Delivery services are probably the biggest obstacle to customer satisfaction with pur— chase decision—making through telecommuting. There are currently three home delivery methods available. Some stores operate their own truck delivery fleet. More often United Parcel Services, who work with retailers on a regu— larly scheduled route basis, or Parcel Post, Railway and Air Express are used. None of these delivery services presently provide the type of quick service which would be desirable in conjunction with telecommuting.77 Hope— fully by the time the public needs it, a more efficient means of delivery will have been developed. Similarities 33 in the socioeconomic status of nonstore retailing cus— tomers and frequent users of delivery services, with the single exception of including the elderly as delivery service users,78 may help to provide incentives to decisionmakers in regard to this need. Banking Conditions Growth of mechanization. Today the demand for financial services in terms of access to savings and checking accounts, credit cards, loans and direct payment plans is high and continuing to grow. Consumers have adopted a convenience orientation. They want these services available and rapidly performed at all times, wherever they may happen to be. People also expect banks to carry out these services with a relatively high degree of protection and security.79 To keep up with the growth in service demands, such as a 657 percent increase (from 3.5 billion to 23 billion) in the number of checks being processed between 1939 and 1973,80 banks have gradually computerized many of their internal operations. Often these automated systems have closely duplicated the previous paper accounting work, but at least the work has been done at a faster rate. As the sheer amount of paper being processed has grown, as well as the labor costs on the remaining employees, it has been acknowledged by many 34 bankers that fuller use of computer technology, although costly, would be more efficient.81 In response to these consumer and business demands, the concept of a “cashless society" has been introduced. The key to offering rapid, paperless financial service transactions at any time of the day or night has been identified as the electronic transfer of funds, often called simply EFT. At this time, EFT through the use of electronic communication produced networks of computers can ideally provide: 1. automated tellers-~which can carry out cash with- drawals, deposits, transfers, balance checks, and payments depending upon the machine set up. 2. automated clearing houses—-providing direct pay— ment of salaries and bills 3. point-of—sale equipment——to verify the status of an account or actually transfer funds between customer and business accounts. Consumer fears hindering the adoption of EFT tech- nology, which may also enter the realm of TPDS, are that consumers are generally afraid of computer mix—ups or the loss of privacy and control over their own affairs.83 Yet the conveniences of using EFT, where it is currently available, appear to be winning out over many of these fears. An initial step in consumer acceptance of EFT is Willingness to let machines handle financial transactions. One trend in banking services which is easing consumer fears, while not quite reaching the full changeover to EFT, is telephone banking. Although this form of banking 90t off to a somewhat slow start,84 its use seems to be 35 picking up, particularly among the young, higher income families and the elderly.85 The use of EFT in general is also growing.86 For example, the Wilmington Saving Fund Society, a bank in Wilmington, Delaware which offers its depositors a fairly comprehensive program (including automatic payroll deposits, bill payment, local retail purchases, as well as the more standard checking and savings account with- drawals, deposits, and transfers) reported that between 1973 when the program was started and 1975, while the total number of items processed by the bank increased by 138 percent, EFT items increased 680 percent.87 The user of EFT has been described as younger, more highly edu- cated and affluent.88 However, avid users have been found across groups of college students, housewives, minorities, and married men.89 A financial question posed by telecommuting tech— nology is whether or not buyers should be allowed to direct the flow of funds out of their accounts into the accounts of sellers from in the home. Given the offers of direct bill payment and point—of—sale transactions, being made by banks where they are allowed to do so, it does not seem unreasonable to assume that banks might be in favor of having EFT operate in the future in con- junction with purchase decisionmaking via telecommuting. 36 The Service System As stated earlier, a nationwide network of two-way CATV systems would be needed in order to achieve maximum economies of scale; and system growth will depend upon the demand for its services. TPDS, as a potential service to be offered, would require a polling system which could give individual homes access to the service source. This source would be linked to the computerized warehousing systems of participating merchants. Consumers could then transmit requests for product information and, consequently, purchase orders to these warehouse operations. When neces— sary, people could be made available for full audio-video communications, but the majority of product information would be delivered via computer. If EFT and home delivery of goods were added to this service, the entire purchase decisionmaking activity could be performed without ever leaving the home. The question at hand is whether people would adopt TPDS as a viable means of making purchase decisions to an extent which would promote service growth. Expansion in the nonstore retailing industry and in the use of self— service store designs, along with the success of several communication media-oriented shopping systems, indicates that there are people who are willing to shop without much personal interaction with store employees. The increasing use of EFT in conducting banking transactions —'———r a 37 complements these trends by displaying a willingness to conduct banking transactions through the use of mechanical or electronic devices. If acceptance of these retailing and banking related behaviors continues to grow, TPDS, as an energy efficient means of performing such behavior, could be in high demand. Given that purchase decision— making, in itself, is an activity which is frequently conducted by most people, if people had the desire to make a majority of their purchase decisions via TPDS, the necessary system momentum could occur. Limitations and Implications In order to assess the extent to which the above chain of events might occur, one must determine just how favorable certain types of people are likely to be towards TPDS. Since there has not yet been any full—scale oper- ations of such a system, one cannot study human factors relative to actual usage levels. In fact, given the lack of literature concerning consumers‘ perceptions of tele- commuting, especially in—the—home services, no firm basis exists for selecting factors for study. Therefore, this study is exploratory in nature. Following the development of a list of potentially relevant human factors from research conducted in a number of fields, projective techniques have been used to estimate how favorable individuals might be towards the concept of TPDS. These individuals have been classified on the basis of the —7— 38 human factor characteristics, and then an attempt has been made to identify relationships between the varying levels of favorability and the human factors. This process has operated under two major limi— tations. First, despite an extensive literature search, the human factors selected as potentially relevant may not be truly appropriate to the TPDS situation. Second, the concept of TPDS may be difficult for some subjects to grasp, thereby biasing the favorability measures. Hope— fully pretesting of the favorability items will reduce this source of error. The findings of this study should be relevant to policymakers in several fields, particularly mass media, urban planning and social psychology. TPDS usage, alone, would necessitate changes in the retailing, banking and advertising industries. Store and bank operations could become much more automated or electronically conducted. Full utilization of the interactive cable medium could greatly alter advertising techniques. If telecommuting were extended to many other services beyond purchase decisionmaking, such as work or education, the histori- cally limiting factors of transportation and communication would experience a greater reduction than ever before. If people were capable of reaching whatever they need from their home, the selection of home sites could become much more self—oriented than it is now. That is, 39 such decisions could become more dependent on social relationships than work obligations. Greater flexibility in the scheduling of daily activities could also be achieved. With this, people might be able to plan their less desirable "chores" around their more highly valued activities rather than the other way around. In the long run, telecommuting thus contains properties which could allow a higher level of social interaction than is currently enjoyed. Going beyond the question of how telecommuting may affect individuals when first intro- duced, there is then the question of whether these long— range social attainments are to be preferred over what society currently experiences. It is possible that telecommuting may not ever be made available to the American public. However, given the current energy conservation problems of the nation, it is likely that TPDS or something similar to it will eventually be needed. In either case, it will be impor- tant for policymakers to know what factors may affect the situation and what those effects may be. This study is designed to begin this identification process. Presentation of this study is broken up into five sections or chapters plus three appendices. This first chapter has sought to introduce the problem to be explored and define most of the fundamental aspects of that problem. The second chapter pursues behavioral 4O theory related to the TPDS situation. Interrelationships found between the conceptual, situational and behavioral factors are used to derive a statement of the independent and dependent variables and the specific hypotheses to be tested. The third chapter describes all of the data col- lection and treatment procedures along with the statisti— cal methods used to test the hypotheses. The fourth chapter reveals the findings of the hypotheses testing, while the fifth chapter contains the conclusions drawn from the findings, including recommendations for future research. The telephone questionnaire used for data collection will be found in Appendix A. Appendix B con— tains the interitem and intervariable correlation matrices. The multiple regression statistics are displayed in Appendix C. CHAPTER I--NOTES lDouglas Jones, ed., Communication and Energy in Changing Urban Environments (London: Butterworth and Co., Ltd., 1970). 2Ithiel de Sola Pool, ed., Talking Back: Citizen Feedback and Cable Technology (Cambridge: The MIT Press, 1973); Ben Bagdikian, The Information Machines (New York: Harper and Row, 1971); A. J. Dakin, "Introductory Address," in Communication and Energyfiin Changing Urban Environments, ed. Douglas Jones (London: Butterworth and Co., Ltd., 1970); IEEE Transactions on Communications, Vol. COM-23, No. 10 (October 1975), entire issue. 3Lesley A. Albertson, "Telecommunications as a Travel Substitute: Some Psychological, Organizational and Social Aspects," Journal of Communication (Spring 1977): 33; Richard C. Harkness, Telecommunications Substitutes for Travel, Special Report to the Office of Telecommuni- cations, U.S. Department of Commerce (OT—SP—73-2, COM-74— 10075, December 1973). 4Albertson, "Telecommunications as a Travel Sub— Stitute." 5"Technology as a Deterrent to Dehumanization," Science 185 (August 30, 1974): 1. 6Alex Reid, New Directions in Telecommunications Research (a report prepared for the Sloan Commission on Cable Communications, 1971), p. 3. 7Natan Katzman, "The Impact of Communications Technology: Promises and Prospects," Journal of Communi— cation 24 (Autumn 1974): 47-58. 8Francesco Nicosia, Consumer Decision Processes (Englewood Cliffs, N.J.: Prentice-Hall, 1966); Alan R. 41 42 Andreasen, "Attitudes and Consumer Behavior: A Decision Model," in New Research in Marketing, ed. Lee E. Preston (Berkeley: University of California Institute of Business and Economic Research, 1965), pp. l-l6; William F. Massy, David B. Montgomery, and Donald G. Morrison, Stochastic Models of Buying Behavior (Cambridge, Mass.: The MIT Press, 1970); John A. Howard and Jagdish N. Sheth, The Theory of Buyer Behavior (New York: John Wiley and Sons, Inc., 1969); R. Lavidge and G. A. Steiner, “A Model for Predictive Measurements of Advertising Effectiveness," Journal of Marketing 25 (1961): 59-62; Everett Rogers and F. Floyd Shoemaker, Communication of Innovations (New York: The Free Press, 1971), pp. 177-85. 9Nels Anderson and Edward C. Lindeman, Urban Sociology (New York: Alfred A. Knopf, Inc., 1928), pp. 43— 49, 146-53, 204-19; Harold M. Mayer, "Urban Geography and City and Metropolitan Planning," in Social Science and the City: A Survey of Urban Research, ed. Leo F. Schnore (New York: Frederick A. Praeger, 1968), pp. 221-38; J. Langdon, "Human Interactions and Communications Technology: A Two—Way Process," in Communication and Energy in Changing Urban Environments, ed. Douglas Jones (London: Butterworth & Co., Ltd., 1970), pp. 78-82. 10A. J. Dakin, "Introductory Address," in Communi- cation and Energy in Changing Urban Environments, ed. Douglas Jones (London: Butterworth & Co., Ltd., 1970), p. 12. llIbid., p. 13. lZIbid. 13Ibid., p. 12. l4Melvin M. Webber, "Transportation Planning for the Metropolis," in Social Science and the City: A Survey of Urban Research, ed. Leo F. Schnore (New York: Frederick A. Praeger, 1968), p. 390. 15Robert B. Mitchell and Chester Rapkin, Urban Traffic: A Function of Land Use (New York: Columbia University Press, 1954); Dr. J. Douglas Carroll,"Jr., "Detroit Metropolitan Area Transportation Study, and Alan M. Voorhees, "The General Theory of Traffic Move- ment," cited by Webber, "Metropolis," p. 396; John Rannels, 43 The Core of the Cit : A Pilot Study of Chan in Central Business Districts New York- 0 um 1a Uni— versity Press, 1956); Webber, "Metropolis," p. 396. l6Webber, "Metropolis," p. 397. 17John F. Kain, "Urban Travel Behavior," in Social Science and the City: A Survey of Urban Research, ed. Leo F. Schnore (New York: Frederick A. Praeger, 1968), p. 164. 1 8"Energy: Ford Foundation Study Urges Action on Conservation," Science 186 (November 1, 1974): 426. 9"Nuclear Moratorium: Study Claims that Effects Would Be Modest, Foresees Low Growth Rate for Total Energy Demand," Science 195 (January 14, 1977): 156-57. 20"Academy Study Finds Low Energy Growth Won't Be Painful," Science 195 (January 28, 1977): 380. ZlIbid. 22Bureau of the Census, U.S. Department of Com— merce, Statistical Abstract of the United States 1976, Table No. 893: Newspapers and Periodicals: 1950 to 1976, p. 540. 23Ibid., Table No. 875: Telephone and Telegraph Systems: 1950 to 1974, p. 532. 24Ibid., Table No. 893: Newspapers and Periodi— cals: 1950 to 1976, p. 540. 25Ibid., Table No. 882: Commercial Broadcast Stations, Number and Revenues: 1950 to 1974, p. 535. 26Ibid. 27Ibid., Table No. 884: Cable Television-~Systems and Subscribers: 1952 to 1975, p. 536. 28Ibid., Table No. 875: Telephone and Telegraph Systems: 1950 to 1974, p. 532. 44 29Ibid., Table No. 877: Telephones, Calls, and Rates: 1950 to 1975, p. 533. 3OIbid., Table No. 873: Volume of Mail, by Classes: 1960 to 1975, p. 531. 31Ibid., Table No. 986: Motor-Vehicle Travel, by Type of Vehicle: 1940 to 1974, p. 592. 32Ibid., Table No. 912: Consumption of Fuel Resources, by Major Consumer Group: 1970 to 1975, p. 551. 33Based on a telephone conversation with Mr. Mitchell, Operating Superintendent, Sears Roebuck & Company, Lansing, Michigan. 34The major factor contributing to this assumption is the fact that cable television, by being potentially able to simultaneously carry more information than other forms of wave transmission, is more efficient. Given that the bandwidth or speed of wave transmission determines what kind and how many bits of information may be trans- mitted per unit of time, the efficiency of a communication mode can be attributed to the relative size of its band- width. The magnitude of cable television efficiency is reflected in the fact that while the transmission band— width available through a telephone system is 3.1 kilo- cycles per second, AM radio 6 kilocycles per second, FM radio 15 kilocycles per second, and over-the-air television 6 megacycles per second, cable television offers 300 mega- cycles per second. This means that, given messages of an equal form over a specific period of time, cable tele- vision has the capacity for carrying 50 times more mes- sages than regular television and 97,000 times more mes- sages than a telephone. 35Ben H. Bagdikian, The Information Machines (New York: Harper and Row, 1971). 36Ithiel de Sola Pool, Charles Murray, and Kenneth Dobb, "Prospects for On-Demand Media," in Talking Back: Citizen Feedback and Cable Technology, ed. Ithiel de Sola Pool (Cambridge: The MIT Press, 1973), p. 266. 37Ibid. 45 38Walter S. Baer, Cable Television: A Handbook for Decisionmaking (New York: Crame, Russak & Company, Inc., 1974), p. 3. 39Martin P. Block, James B. Wright, and D. Stevens McVoy, "An Evolutionary Approach to the Development of Two-Way Cable Technology Communication" (publication forth— coming in IEEE Transactions on Cgmmunications). 4ORadin and Smith, "Technologies," p. 13. 41Ibid., p. 5. 42Baer, Handbook, pp. 7—9. 43The Sloan Commission, On the Cable--The Tele- vision of Abundance (New York: McGraw—Hill Book Co., 1971), p. 42. 44Pool, Murray, and Dobb, "Prospects," p. 266; Block, Wright, and Stevens, "Evolutionary Approach"; Bagdikian, Machines. 45Paul Baran, "Broad-Band Interactive Communica- tion Services to the Home: Part II—-Impasse," IEEE Trans- actions on Communications, COM—23, No. 1, pp. 178—84; Block, Wright, and Stevens, “Evolutionary Approach"; Pool, Murray, and Dobb, "Prospects," pp. 266-316. 46 Baran, "Impasse,' pp. 178-79. 47Anne W. Branscomb, "The Cable Fable: Will It Come True?" Journal of Communication (Winter 1975): 50-62. 48Pool, Murray, and Dobb, "Prospects," p. 267. 49Block, Wright, and Stevens, "Evolutionary Approach," p. 3. 50Ibid., pp. 3-6. SlIbid., pp. 7—10. 52Ibid., pp. 3—4. 46 53Delmer C. Ports, "Trends in Cable TV," IEEE Transactions on Communications, COM—23, No. 1, p. 96. 54Robert D. Briskman and Leonard S. Golding, "Role of Satellites in Broad—Band Interactive Cable Networks," IEEE Transactions on Communications, COM-23, No. 1, pp. 88- 91. 55Paul Baran, "Broad-Band Interactive Communication Services to the Home: Part I-—Potential Market Demand," IEEE Transactions on Communications, COM—23, No. 1, pp. 5— 14; Pool, Murray, and Dobb, "Prospects"; Block, Wright, and Stevens, “Evolutionary Approach"; Hollowell, Handbook 1975—1976; Gerhard J. Hanneman, "Communication Substitutes for Transportation: The Movement of Information Versus the Movement of People" (paper presented at International Communication Association, New Orleans, Louisiana, April 1974). 56Pool, Murray, and Dobb, "Prospects," p. 268. 57Dakin, "Introductory Address," p. 12. 58James D. Scott, Cable Television, Strategy for Penetrating Kev Urban Markets, Michigan Business Report No. 58 (Ann Arbor: University of Michigan, 1976), pp. 61- 62. 59Ibid., pp. 62-64. 6OIbid., pp. 67—71. 61Ibid., pp. 64-67. 62Ibid., pp. 41-61. 63"TV Cable Moves Firefighting Lesson, Data Response," Minicomputer News. 64John Ledingham and Todd Henning, "Cable TV in Columbus: Can It Deliver?" Columbus Monthly, June 1977, pp. 46—51. 65Sally Koris, "Columbus Viewers Will Take Part in Programs When Two—Way Service Begins Late in the Year," The Wall Street Journal, 10 August 1977, p. 32. 47 66Larry D. Redinbaugh, Retailing Management—~A Planning Approach (New York: McGraw Hill Book Company, 1976)] p- 33. 67Ibid., p. 34. 68Roger A. Dickinson, Retail Management—-A Channels Approach (Belmont, Calif.: Wadsworth Publishing Co., Inc., 1974). p. 8. 69Ibid. 70Stuart U. Rich, Shopping Behavior of Department Store Customers (Boston: Harvard University Graduate School of Business Administration, 1963), p. 12. 71Redinbaugh, Planning Approach, p. 39. 72Rich, Department Store Customers, p. 81. 73 Redinbaugh, Planning Approach, pp. 10-14. 74Ibid., pp. 10-32. 75Dickenson, Channels Approach, p. 13; Redinbaugh, l Planning Approach, p. 21. 76Carl M. Larson, Robert E. Weigand, and John S. Wright, Basic Retailing (Englewood Cliffs, N.J.: Prentice- Hall, Inc., 1976), p. 464. 77Ibid., pp. 370—73. 78Rich, Department Store Customers, p. 81. 79Robert D. Richley, "Branch Banks and CBCTS: A Delicate Balance," Bankers Monthly Magazine, March, 1976. 80Jane Rippeteau, "Comment," Finance Magazine, February, 1976. 81"A Retreat from the Cashless Society,‘ Week, April 18, 1977, p. 81. Business 48 82"Autotransaction Industry Report," Autotrans- action, March 3, 1976, p. 4. 83"ATMS vs. Passbooks," American Banker, October 13, 1976, p. 4. 84Morton C. Paulson, "Telephone Service," citing work done by The National Commission on Electronic Fund Transfers and a study conducted by Unidex Corp of Bloom— ington, Indiana (reprint of Newspaper Article--source unknown). 851bid. 86"Planning and Experience in Automated Teller Machines," Bank News, June 15, 1976, pp. 23—25. 87Francis P. Evers, "The WSFS Plan-~A Practical Experiment in EFTS," The Magazine of Bank Administration, April, 1976, pp. 52-55. 88Ibid., p. 53; Linda Fenner Zimmer, "Survey of EFT Equipment Manufacturers" (speech presented at 28th Annual Bank Operations Conference of the National Associ— ation of Mutual Savings Banks, date and location of con- ference not cited). 89Ibid. CHAPTER II LITERATURE REVIEW AND DEVELOPMENT OF THE HYPOTHESES Following Reid's suggestion that communication technology assessment should be based upon a combination of technological feasibility and human factors,1 the literature search was divided into three sections. As discussed in the previous chapter, the role of communi- cations technology in man's environment (conceptual factors) was explored in order to initially assess the need for telecommuting. Next, conditions within the broadband communications, retailing and banking indus— tries were examined (situational factors); resulting in the acceptance of TPDS as a viable form of telecommuting for study and providing some information regarding the potential users of TPDS. In this chapter, the range of human factors which may be associated with favorability towards TPDS will be expanded by considering attitudes towards shopping and teleconferencing (behavioral factors). Chapter II concludes with selection of the research vari— ables and development of the research hypotheses. 49 50 Shopping Versus Buying In determining the potential range of reactions towards TPDS, one must examine the general form of behavior associated with this activity. It is difficult to clearly separate and define the variations in behavior related to all instances of purchase decisionmaking. In this study, the most basic classification of purchase decisionmaking behavior will be considered to be the dis- tinction between buying (which will be defined as decision— making behavior which has ended with a product purchase) and shopping (meaning spending time looking at products, or being in a store without necessarily purchasing any— thing). If the motivations behind going into a store are related only to those which direct buying behavior, the decision to go into a store will occur when a person's need for particular goods becomes sufficiently strong enough for him to allocate time, money, and effort to visit a store.2 In this case, TPDS, by allowing people easy access to products they wish to purchase, might not be that difficult for people to adjust to. Tauber, however, has suggested that people may go to a store just to shop. In an exploratory study of why people go to a store, he found two categories of moti- vations: Personal——role playing, diversion, self—gratification, learning about new trends, phySical actIVity, and sensory stimulation 51 Social-—experiences outside the home, communications With others having a similar interest, peer attraction, status and authority and pleasure of bargaining (when applicable).~l3 If large segments of the U.S. population do indeed go to stores just to shop rather than buy, then telecommuting, by eliminating most in-store stimuli, will not be as likely to be perceived as a motivationally satisfying means of purchase decisionmaking. Memmot has stressed that trips for the purpose of observing activities, objects, or people may be amenable to audio-visual substitution but partici- pation in activities or spending physical time with objects or people are not.4 What this may actually amount to is the preference of people for impersonal versus inter- personal communication oriented transactions. If people are satisfied by the stimuli they receive via telecommut— ing, they may feel no need to go personally to a store. Stone, in a 1954 study of the orientations of Chicago housewives to large chain stores versus smaller local stores, found four approaches to store choice: Economical ——Primarily concerned with buying, a store is judged in terms of product price, quality and variety. A store and its personnel are merely the instruments by which one purchases goods. Demographics: Young, with children, lower social class Personalizing——A close fundamentally interpersonal oriented relationship exists between the customer and store personnel, giv— ing shopping a strong social function. Demographics: middle aged, with few or many children, lower social class, low social participation Ethical 52 ——A moral obligation to shop in smaller, local stores over larger chain stores exists. Store personnel are considered only to the extent that large stores are "killing" small ones. Demographics: older, higher social class, higher social participation Apathetic——Not particularly interested in shopping. Convenience in terms of location or avail- ability of all desired items in one location is the only dominant feature. Demographics: older, lower class, high social participation. These alternative approaches seem to indicate that while interpersonal communication in connection with shopping activities is valued rather highly by some people (per— sonalizing), (economical, others do not view it as a major concern ethical, apathetic). The previous discussion of the retailing industry indicated that, today, nonstore retailing and large self—selection or self—service store designs are generally preferred by consumers over full service store use. These trends seem to represent a dom— inance of economical or apathetic approaches to store choice over the personalizing or ethical orientations. Assuming that an individual having a personalizing orien— tation would TPDS than an tations, the to result in among people experience a greater behavior loss when using individual possessing any of the other orien— retailing industry trends could be expected a higher level of favorability towards TPDS in general. 53 Product Differences The above conclusion is encouraging in terms of promoting the use of TPDS. There are, however, other aspects of the communications-travel exchange which may affect the favorability or impact of TPDS. Hanneman has pointed out that the pleasure aspect of physical transpor— tation probably cannot be replaced, people do like to travel for certain nonroutine functions and for routine functions if the site is appealing.6 It is possible that the routine nature of consumer buying habits in regard to particular kinds of products can be viewed in a similar fashion. Consider the following product classifications. Convenience Goods—-These goods, often called staples, tend to be low in cost. Consumers generally have a thorough knowledge of these products prior to enter- ing a store. Although price may play a determining factor in specific brand selections, this product is usually purchased on a routine and continuous basis. Shopping Goods—~The unit value or price of this type oprroduct is usually greater than that of convenience goods and they are purchased less frequently. Con- sumers, feeling less knowledgeable about these goods, often shop around comparing product features, prices, warranties, and other characteristics. All durable goods fall into this category as well as such things as jewelry, clothing and small appliances. Specialty Goods--The most prominant characteristic of these goods is brand preference. They are low or high priced products about which the consumer has a lot of knowledge. The consumer will accept no substitutes, thus as long as the product is available at a certain store, purchase behavior may be fairly routine. When necessary, the consumer will exert extra effort to find these products. Impulse Goods—~Products of low unit value which con— sumers purchase with little or no forethought fall into this classification. While not preplanned 54 purchases, they may be fairly commonly used items. A major factor in the purchase of impulse goods is the opportunity to see them while in the store. If Hanneman's conclusions are correct, one might expect consumers to be more willing to use telecommuting for the purchasing of convenience goods and possibly specialty goods than shopping goods. The value of tele- commuting to shopping good purchase decisionmaking will depend upon consumers' perceptions of the information pro— vided. If they can learn enough to feel capable of making satisfactory selections between alternative goods, tele- commuting might be more acceptable. The conversion of purchase decisionmaking activity which currently involves some sort of in-store trial, such as trying on clothes, will probably meet with the greatest degree of opposition. The level of impulse product purchase behavior which can occur with telecommuting will depend upon the form of product presentation chosen for use by sellers. If con- sumers can call for information only about specific pro- ducts, the conditions necessary for impulse purchases will be lost. If consumers receive sequential information packages, they will be exposed to more than the products in which they were most interested, and impulse product purchase behavior as we know it today can still occur. Social Participation Another factor which may contribute to travel pleasure is the company of others either during, prior to 55 or immediately following the trip. That is, while inter— personal communications with store personnel may not be a relevant factor, discussion regarding the trip with other people could be. Rich, in a survey of department store customer shopping habits, found that three out of ten housewives shopped with others regularly or fairly often, and five out of ten shopped with others once in a while.8 These figures indicate that the company of others may be important to at least certain segments of the population. Unfortunately, no attempt was made by Rich to identify 1 these segments by demographic characteristics. NOr did A Rich explore the frequency of interpersonal communication experiences surrounding the trip. Of the people who prefer to shop alone, some of them may feel a lesser need for interpersonal communication while in a store because they receive all of this form of communication they need during nonpurchase decisionmaking activities. This type of trade-off could play an increas- ingly important role if telecommuting services are used to conduct a broad range of currently interpersonal communi- cation—oriented activities. Teleconferencing Behavior Teleconferencing research, encompassing the areas of effectiveness of teleconferencing as a general mode of communication between individuals, as a means of conducting various tasks, and as a change in the organizational 56 structure of business operations, suggests that telecommut- ing may not always be favorably perceived as a replacement for all purchase decisionmaking activities. Teleconferencing systems have been started by various companies and governmental operations in a number of countries including Britain, Canada, the United States, Japan and Australia.9 For example, the First National City Bank of New York, the Banker's Trust of New York, and the First National Bank of Boston have a privately owned, two-way audio—video teleconferencing system. It utilizes special rooms which each contain one camera to View conferees and another for documents. All locations are interconnected via coaxial cables leased from AT&T. The bank users report that this system is the best method available to obtain information and render decisions when . . . . 10 quick reactions are necessary in stock market operations. Other systems cover much longer distances such as the British Post Office's Confravision System which connects the cities of Birmingham, Bristol, London and Manchester or Dow Chemical's system linking Midland, Michigan and Freeport, Texas. Dow's system, like the bank system, is perceived to be serving in the company's best interests. The British system, which is available for any business located in the five cities, however, has projected a slow growth potential due to: .- 57 1. British business management has an inherent con— servation to innovation. ‘ 2. Distances between the system's studios are rela— tively short. 3. The five studios cover only a limited geographical area. 4. Many businessmen in the United Kingdom are reluc— tant to forego the personal advantages of travel. 5. The apparent inability of the system to meet the user's seeming need for personal contact at every meeting. 6. The inability of the system's managers to define precisely the cost/benefit relationships that exist between use of the system and travel, particularly as those relationships affect users' operations effi- ciency. The major complaint of teleconferencing users regarding communications with others is that it is more difficult to communicate with people they are not very well acquainted with.12 Albertson suggests that this is due to the loss of the tactile, gustatory and olfactory dimensions which occurs when one can no longer shake hands or perhaps share a meal with business associates.13 Essentially, as discussed by Conrath and Thompson, a feeling of discontinuity results from one's inability to share environments.l4 If Albertson's observations are correct, this problem might not be as strongly associated with the TPDS situation because salesperson- customer exchanges generally involve less personal activities. In studying the ability of people to perform a range of simple to complex tasks over different media forms, the Communications Studies Group at London Uni- versity found that the choice of medium (between telephone, r______f closed circuit TV and face—to—face) did not affect the [ outcome of simple tasks such as information exchange.15 Albertson found, however, that a visual image may actually sometimes serve as a distractor; telephone conversations resulted in better information recall than face-to—face or videophone exchanges.l6 Thus, in the case of purchase decisions that are repeat purchases or where people know precisely what they want, video displays of products on the downstream channel and complete audio—video ordering capabilities might not be necessary. Yet since it is not likely that all consumers would be in this position for the same products at the same time, video messages or at least a chance to choose between using or not using video would probably be needed. The use of both audio only and audio-video media in conducting more complex tasks, such as communication involving persuasion or negotiation has been rated as less satisfactory than face-to—face communications by teleconference users.17 This implies that the personal selling approach often found in connection with shopping goods could probably not be effectively transferred directly to audio-video exchanges. The most notable change in the organizational structure of businesses Choosing to use teleconferencing on a full—scale basis is the spreading out of one's work force from offices to homes. This diffusion of efforts 59 is said to result in the loss of managerial input "essen- "18 and tial for providing direction, motivation and morale the social interaction "by which people enhance their self esteem as persons of competence, and satisfy their needs 19 This for companionship, affiliation and belongingness." social interaction factor can be considered comparable to the problems TPDS may present to the personalizing shopper. As stated earlier, the loss of social interaction in the teleconferencing situation may perhaps be replaced by interaction obtained during other activities. Yet with more activities being converted to telecommuting, the supply of social need satisfaction could dwindle. Another issue raised by Albertson, drawing upon the work of Argyle and Little, is the value of keeping the role behaviors of work ("functional expertise and imper- sonal detached attitudes“)20 and homelife ("warmer, more relaxed relationships")21 separate. He states that the blending of these behaviors by teleconferencing could cause people to lose touch with reality or psychological stress.22 Such an effect could apply to TPDS particularly among those people who define themselves via the role of providing food (grocery shopping) or other products for their household. The extent to which TPDS will be perceived as discussed in regard to these teleconferencing findings will undoubtedly vary from person to person. The one 60 attempt made to actually predict preferences for telecon- ferencing over physical travel to face—to-face exchanges on the basis of such considerations as attitudes towards travel and the perceived unsuitability of teleconferencing for particular types of meetings failed to find any “cor- relations large enough to make such predictions." The researchers concluded that "either the decision to sub— stitute may be an idiosyncratic one which the present study did not tap," or that "conceptualizing substitution of telecommunications as a replacement process rather 23 If the than a supplement to travel is ill—conceived." second explanation is true, TPDS could never fully replace the form of buying and shopping activities that exists today. General Classification of Reactions towards TPDS If reactions can be predicted to TPDS, it is cer- tain that one cannot expect to find a single standard reaction among the public towards TPDS. Following the conclusions of Stone, Tauber and teleconferencing research, there are basically two types of reactions which might occur: Generally Favorable in that one can buy products -—under very convenient circumstances ——very efficiently in terms of comparing price, quality and variety --via impersonal communications, avoiding con- tact with store personnel 61 Generally Unfavorable because -—much of the in—store shopping related stimuli are lost ——interpersonal communications is preferred —-dislike of innovation or change, particularly where bigger operations are replacing small ones Chapin, in his work concerning the behaviors of population aggregates in relation to urban planning, has developed a general systems model which appears appropriate to the problem of determining the human factors associated with these reactions. This model (shown in Figure 1) con— siders factors which, within a selected temporal-spatial framework, may affect an individual's level of partici— pation in a particular activity.24 Chapin attributes any decision made to the joint action of two factors—~propensity to engage and opportunity to engage. Propensity to engage is defined by Chapin as consisting of the interaction between motivations and thoughtways (meaning wants and attitudes) which, as ener- gizing factors, predispose individuals towards action, and roles and personal characteristics (referring to demographics) which, as constraining factors, precon— dition actions.25 Opportunity to engage is considered to be depen— dent upon the interaction of individuals' perceptions regarding their ability to have access to necessary facil- ities or services and the quality of those facilities or services. Relating the opportunity concept back to that of propensity, Chapin assumes that in order for an 62 Motivations & Thoughtways Predisposing Action Roles and Person Characteristics Preconditioning Action Prop Sity to Engage Activity Patterns Perceived Availability Opp tunity to Engage . . . .—_______9 of FaCility or Serv1ce Perceived Quality . . . ——_______9 of FaCility or Serv1ce Fig. 1. Chapin model SOURCE: F. Stuart Chapin, Jr., Human Activity Patterns in the City: Things People Do in Time and in Space (New York: John WIley & Sons, Inc., 1974), p. 33. —: i 63 activity to occur, access to the related facilities or services must be perceived as of a high enough quality, and sufficiently convenient to permit the level of use desired through the propensity to engage factor.26 This model is ideal for work related to the adoption of a new technology because it thus accounts for attitudes towards the technology and related behavior, demographic characteristics which may affect the adoption decisionmaking process and perceived availability and quality of the technology. Unfortunately, since TPDS is not actually available to consumers at this time, per- ceptions of TPDS availability and quality cannot be assessed. However, it is believed that certain behavioral factors can be substituted for the TPDS availability and quality mea5ures because peoples' perceptions of a new activity, in terms of the effort involved to gain access (availability) and performance standards (quality), may be affected by their past levels of participation in activities having certain characteristics in common with the new activity. Diffusion of innovation research find- ings suggest that the degree to which an innovation is perceived as being compatible with past experiences, while offering more than the idea, good or service it super- sedes, affects an individual's perception of that inno— vation.27 Therefore, the frequency with which an indi— vidual conducts impersonal versus interpersonal oriented 64 shopping and banking transactions will be used as indi- cators of future perceptions of TPDS availability and quality. Selection of Independent and Dependent Variables Pulling together all of the demographic character— istics which have been attributed to people practicing or not practicing behavior similar to that which may be involved in TPDS, one finds the patterns displayed in Table 3. In examining these conclusions, one finds several contradictions in regard to work done by Stone. These contradictions could be the result of the orien— tation or timing of Stone's work. They might also be present because the behaviors of interest cannot be rep— resented by orthogonal demographic descriptions. Whatever the reason may be, Stone's findings were ignored during hypothesis construction. On the basis of the above con— clusions, the independent demographic variables examined were: Sex.-—A distinction is made between males and females. Socioeconomic Status.--A socioeconomic index rating based upon the occupation of the highest house- hold wage earner. Working Women.——This classification requires a min— imum of ten hours per week of work outside of one's home. SUMMARY OF THE LITERATURE: MAY AFFECT REACTIONS 65 TABLE 3 ATTRIBUTES WHICH TOWARDS TPDS General Favorability towards TPDS Source Favorable Unfavorable (High Impersonal (High Interpersonal Orientation) Orientation) Katzman More educated Less educated Higher socioeconomic Lower socioeconomic status status More mass media Less mass media exposure exposure More social partic- Less social partici- ipation pation Greater physical Less physical mobility mobility Nonstore More educated retailing Higher socioeconomic including status delivery Women with children services Elderly Less mobile Telephone Higher socioeconomic banking status Young and elderly EFT More educated Higher socioeconomic status Young Stone Lower socioeconomic Higher socioeconomic status Young women with children Higher social par- ticipation in the community status Lower socioeconomic status Middle aged w/many or few children Elderly Higher social partic- ipation in the com— munity Lower social partic- ipation in the com— munity 66 Educational Status.--Determined by the level of the last year of school attended. Lifecycle Status.-—People are identified on the basis of their marital status and whether or not they have children in their household. Age.——The major distinction is between younger and older adults. Social Participation.——The frequency of partici- pation in a broad range of activities. Mass Media Exposure.-- Broadcast.—-The amount of time people spend watching TV and/or listening to the radio. Egip§.-—The amount of time people spend reading newspapers and/or how many magazines they read on a regular basis. ggplg.-—A distinction is made between people who currently subscribe or have previously subscribed to cable television and those people who have never sub— scribed. Physical Mobility Status.—-A measurement of the radius of the area travelled around one's home within a one—week period. Cars.——The number of cars in one's household. All of the variables except 3, 8-10 and 12 were drawn directly from Table 3. The working women variable was 67 included to identify a possible bias in the socioeconomic status measure as well as to represent a potentially dif- ferent use of time. Given that a higher socioeconomic status and a greater time constraint could occur in con— nection with a working woman, a working woman would be expected to be more favorable towards TPDS than a non- working woman. Mass media exposure is a special case of social participation. Since people who display a high degree of broadcast media exposure have been shown to differ from individuals who prefer print media use,28 these media forms will be measured separately. Experience with cable television will also be measured as a separate variable because of its relevance to TPDS. It was assumed that higher exposure to broadcast or print media and experience with cable television would be associated with more favorability towards TPDS. The number of cars in household variable is included to assess a second aspect of physical mobility, potential availability of cars for travel. The greater the number of cars the less favorable an individual was expected to be towards TPDS. In an attempt to account for the various conceptual and situational factors presented in this chapter, attention was also given to a group of nondemographic independent variables. Two variables sought consideration of attitudes hich might affect one's propensity to engage: 68 Attitudes Towards Technology.-—How people feel about the effects of technology on society. Attitudes Towards Interpersonal versus Impersonal Oriented Shopping Transactions.--A distinction is made between personalizing and nonpersonalizing shopping orien- tations. The remaining variables concentrate on current banking and shopping behaviors which might affect perceptions of TPDS availability and quality: Impersonal Banking Behavior.—-How often subjects use phone, mail and mechanical banking services. Interpersonal Banking Behavior.--How often sub— jects use in—the-bank or drive-in banking facilities. Impersonal Shopping Behavior.-—How often subjects use phone or mail order services. Interpersonal Shopping Behavior.-—How often sub- jects travel to stores by car and secondary measures re— garding when such trips usually occur-—weekend vs. week- days, mornings vs. afternoons or at night. All of these independent variables were to be tested against the concept "favorability towards TPDS." Previous research by Nilles, Carlson, Gray and Hanneman suggested that the operationalization of "favorability" should involve more than a single dependent variable. 69 Nilles et a1. examined two types of attitudes towards a group of eleven telecommuting services: 1. Desirability (the amount of interest people expressed in a service) 2. Value (the amount of money people were willing to pay in addition to a cable TV charge for a service)29 Nilles et al. found that when the services were placed into rank orders based upon their relative levels of desirability or value, the rankings of services across these components tended to vary. In fact, if one calcu— lates a Spearman rank order correlation coefficient for the Los Angeles sample data presented in the study, the relationship between the rankings is shown to be only .36.30 This variance seemed to indicate that the ability of the subjects to project how often they might want to use TPDS could be affected by their perceptions of how much TPDS might cost. Therefore, a decision was made to use two dependent variables: Usage.—-A projected frequency of TPDS use if the service was available for free. Value.-—A projection of how much one would be willing to pay to use TPDS. The value variable was intended, as before, to be an attitudinal measure while the usage variable, as a pro- jection of actual use, was considered to be a behavioral measure . 70 All of the research variables can be perceived to be operating within the environment in many different ways. Using Chapin's model, the independent variables were viewed as fitting into the model as seen in Figure 2. The attitudinal and demographic independent variables are viewed as contributing to propensity to engage and the behavioral independent variables constitute the opportunity to engage factor. All of the independent variables will be tested against each of the favorability components, usage and value. The Research Hypotheses If Chapin‘s model applies to the TPDS situation, one would expect to find three levels of interaction occurring in the TPDS situation: (1) Between each dependent variable and each individual independent variable; (2) Between each dependent variable and each of the two independent variable sets—-propensity to engage and opportunity to engage; (3) Between each dependent variable and all of the independent variables simultaneously. Research hypotheses were developed to test all of these expectations. At the individual independent variable level, research hypotheses were first developed for the motivation 71 General attitudes towards technology and interpersonal 3 versus impersonal shopping transactions Sex I___; Socioeconomic Status————_, Working Women———————————+ Educational Status———————+ Lifecycle Status————————9 Age I‘_e Social Participation_—___+ Mass Media Exposure——————+ Physical Mobility Status—+ Pro ensity to Engage No. of Cars in Household—9 Favorability towards TPDS as reflected by Usage or Value of TPDS O ortunity to Engage Perceived availability and quality of inter- personal verpus impersonal transactions as reflected in performance of inter- personal and impersonal banking and shopping transactions Fig. 2. Data model based on Chapin's work *These transactions always being related to shopping or banking. 72 and thoughtways component of the model, that is, the atti— tudinal independent variables. If favorability towards TPDS is associated with favorability towards technology in general, a positive relationship would be expected to exist between attitudes towards technology and TPDS favora- bility. At the same time, a positive relationship would be expected to exist between TPDS favorability and positive attitudes towards the type of communication provided by TPDS or negative attitudes towards the type of communi— cation not available during TPDS use. The research hypotheses developed to evaluate these expectations were: Hypothesis 1: People having more positive attitudes towards tech- nology will be more favorable towards TPDS than those having more negative attitudes. Hypothesis 2: People having more positive attitudes towards imper— sonal shopping transactions will be more favorable towards TPDS than those having more negative attitudes. Hypothesis 3: People having more negative attitudes towards inter— personal shopping transactions will be more favorable towards TPDS than those having more pOSitive attitudes. Next, the demographic variables, representing roles 1nd person characteristics, were considered. USing the 73 conclusions drawn earlier in this chapter, one would expect to find more favorability towards TPDS amOng: (1) Women, especially those who work over ten hours per week outside of the home; (2 V People who have been or who currently subscribe to cable television; (3 V People from households where there are fewer cars available; (4) People of a (an) (a) (b V (C) (d) (e) (f v (9) higher socioeconomic status higher educational status earlier lifecycle status younger age higher social participation status higher broadcast and/or print media exposure status lower physical mobility status The following research hypotheses were developed to test these expectations: Hypothesis 4: Women will be more favorable towards TPDS than men. Hypothesis 5: People having a higher socioeconomic status will be more favorable towards TPDS than those having a lower socioeconomic status. 74 Hypothesis 6: Working women will be more favorable towards TPDS than women who do not work. Hypothesis 7: People having a higher educational status will be more favorable towards TPDS than those having a lower edu— cational status. Hypothesis 8: People having an early lifecycle status will be more favorable towards TPDS than those having a later lifecycle status. Hypothesis 9: Younger people will be more favorable towards TPDS than older people. Hypothesis 10: People having a higher social participation status will be more favorable towards TPDS than those having a lower level of social participation. Hypothesis 11: People having a higher level of broadcast media exposure will be more favorable towards TPDS than those having a lower level of broadcast media exposure. Hypothesis 12: People having a higher level of print media exposure will be more favorable towards TPDS than those having a lower level of print media exposure. Hypothesis 13: People having some experience with cable television will be more favorable towards TPDS than those having no experience with cable television. 75 Hypothesis 14: People having a lower physical mobility status will be more favorable towards TPDS than those having a higher physical mobility status. Hypothesis 15: People having fewer cars in their household will be more favorable towards TPDS than those having more cars in their households. The behavioral independent variables, as explained previously, were treated as representing participation in activities which might influence perceptions of TPDS availability and quality. That is, it was speculated that an individual who frequently conducts transactions having characteristics in common with TPDS use would be most likely to have positive perceptions of TPDS availability and quality. If perceptions of TPDS availability and quality are affected by levels of participation in these activities, then a positive relationship would be expected to exist between TPDS and the impersonal oriented activi- ties; and a negative relationship would then be expected to exist between TPDS and the interpersonal oriented activities. The research hypotheses developed to cover these situations were: Hypothesis 16: People who conduct more impersonal banking trans- actions will be more favorable towards TPDS than those who conduct fewer impersonal banking transactions. 76 Hypothesis 17: People who conduct fewer interpersonal banking trans- actions will be more favorable towards TPDS than those who conduct more interpersonal banking transactions. Hypothesis 18: People who conduct more impersonal shopping trans— actions will be more favorable towards TPDS than those who conduct fewer impersonal shopping transactions. Hypothesis 19: People who conduct fewer interpersonal shopping trans— actions will be more favorable towards TPDS than those who conduct more interpersonal shopping transactions. At the subgroup level, the propensity to engage factor can be assessed by examining the combined efforts >f the attitudinal and demographic variables. Given the 'elationships suggested in the bivariate level hypotheses, . higher propensity to engage would be defined as involving: (1) (2) (3) (4) People having more positive attitudes towards technology, and impersonal transactions; People having more negative attitudes towards interpersonal transactions; Women who work; People having . . . (a) higher socioeconomic status (b) higher educational status (c) early lifecycle status 77 (d) a younger age (e) higher mass media exposure (f) experience with cable television (9) lower physical mobility status (h) fewer cars in their household Since the model suggests a positive relationship between propensity to engage and favorability towards TPDS, the research hypothesis developed to test this relationship was: Hypothesis 20: People having a high propensity to engage will be more favorable towards TPDS than those having a low propensity to engage. The opportunity to engage factor, as the second subgroup, is comprised of the behavioral independent variables. Again a positive relationship with favora— bility towards TPDS is predicted in the model. Having defined a higher opportunity to engage as involving people who conduct more impersonal or less interpersonal oriented shopping and banking transactions, the research hypothesis constructed was: Hypothesis 21: People having a high opportunity to engage will be more favorable towards TPDS than those having a low opportunity to engage. pensitj model then OI betweer dent ve propens earlier relatic Hyr * Pec eng hav sent a TPDS Us aCcepte factors mEans b in the 78 At the third level, if the joint forces of pro— pensity and opportunity to engage operate as shown in the model (assuming that these forces are properly defined), then one would expect to find a positive relationship between favorability towards TPDS and all of the indepen- dent variables operating together. Defining a higher propensity and a higher opportunity to engage as discussed earlier, the research hypothesis developed to test this relationship was: Hypothesis 22: People having a high propensity and opportunity to engage will be more favorable towards TPDS than those having a low propensity and opportunity to engage. Together, the preceding research hypotheses repre— sent a test of Chapin's model as it relates to the activity TPDS use. Therefore, if these hypotheses fail to be accepted, not only may the relevance of particular human factors be disproved, but also that of the model. The means by which these hypotheses were tested are described in the next chapter. Resear Cable ( 0f Marl cation: 1963, ( p. 20. ObserVe m Travel and Soc 32~43 fof Tre Cations 10075, Suhstit W U.S‘ DE CHAPTER II--NOTES 1 . . . . Alex Reid, New Directions in Telecommunications Research, a report prepared for the Sloan Commission on Cable Communications 1971, p. 3. 2 Edward M. Tauber, "Why Do People Shop?" Journal of Marketing 36 (October 1972): 48. 3Ibid., pp. 47—48. 4F. W. Memmot, "The Substitutability of Communi— cations for Transportation," Traffic Engineering, February, 1963, cited by Hanneman, "Communication Substitutes," p. 20. 5G. Stone, "City Shoppers and Urban Identification: Observations on the Social Psychology of City Life," The American Journal of Sociology 60 (July 1954): 37-45. 6Hanneman, "Communication Substitutes," p. 21. 7Redinbaugh, Planning Approach, pp. 61-66. 8Rich, Department Store Customers, p. 74. 9Lesley A. Albertson, "Telecommunications as a Travel Substitute: Some Psychological, Organizational and Social Aspects," Journal of Communication (Spring 1977): 32-43; Richard C. Harkness, Telecommunications Substitutes for Travel, Special Report to the Office of Telecommuni- cations, U.S. Department of Commerce (OT-SP-73—2, COM—74- 10075, Dec 1973); Charles E. Lathey, Telecommunications Substitutability for Travel: An Energy Conservation Potential, Report to the Office of Telecommunications, U.S. Department of Commerce (OT Report 75—58, January 1975). 79 Telecor Post 01 Albert: M. S. I tems," for the 1973, c "The Be Study, ning Re graphed Technol 921199. communi May 197 Plannin. by Albe: Cation 1 Melbourr Communic and Inte IndUStri Of PS 01‘ Communic BHSineSs Cited by 80 10 . . . Lathey, Substitutability for Travel, p. 91. 11Ibid., p. 79. 12 . . ' L. A. Albertson, "A Preliminary Report on the Teleconference User Opinion Questionnaire," Australian Post Office, Melbourne, 1974 (mimeographed), cited by Albertson, "Telecommunications"; A. E. Casey-Stahmer and M. S. Havron, "Planning Research in Teleconference Sys- tems," Report by Human Sciences Research, Inc., Virginia, for the Department of Communication, Ottawa, September 1973, cited by Albertson, "Telecommunications"; E. Williams, "The Bell Canada Conference Television System: A Case Study, Communications Studies Group, Joint Unit for Plan— ning Research, University College, London, 1973 (mimeo- graphed), cited by Albertson, "Telecommunications." l3Albertson, "Telecommunications," p. 34. 14D. W. Conrath and G. B. Thompson, "Communications Technology: A Societal Perspective," Journal of Communi- cation (Summer 1973): 47—63, cited by Albertson, "Tele- communications." 15Communications Studies Group, "Interim Report, May 1972,‘I Communications Studies Group, Joint Unit for Planning Research, University College, London, 1972, Cited by Albertson, "Telecommunications." 16L. A. Albertson, "A Comparative Study of Communi- cation Effectiveness Across Media," Australian Post Office, Melbourne, 1973 (mimeographed), cited by Albertson, "Tele— communications." 17I. E. Morley and G. M. Stephenson, “Interpersonal and Interparty Exchange: A Laboratory Simulation of an Industrial Negotiation at the Plant Level," British qurnal of Ps cholog 60 (1969): 543—45, Cited by Albertson, Tele— communications." 18E. B. Carne, "Telecommunications: Its Impact on Business," ngyard Business Review 50 (1972): 925—33, cited by Albertson, "Telecommunications." (Paper Psycho commun worth: cation Traits of Soc: "Telec< Travel chology Congres "Telecc in the EBHH‘WI 81 19 L. V. Harvey, "Interpersonal Communication" (paper presented at the annual conference of the Australian Psychological Society, 1971), cited by Albertson, "Tele- communications." 2 .. . 0A. C. Zijderveld, The Abstract Soc1ety (Harmonds— worth: Penguin, 1973), cited by Albertson, "Telecommuni— cations." ’ 21M. Argyle and B. R. Little, "Do Personality Traits Apply to Social Behavior?" Journal for the Theory of Social Behavior 2 (1972): 2—32, cited by Albertson, "Telecommunications." 22Albertson, "Telecommunications," pp. 37-38. 23J. H. Kollen and J. Garwood, "The Replacement of Travel by Telecommunications" (paper presented at the Psy— chology and Telecommunications Symposium, 18th International Congress of Applied Psychology, 1974), cited by Albertson, "Telecommunications." 24F. Stuart Chapin, Jr., Human Activity Patterns in the City, Things People Do in Time and Space (New York: John Wiley & Sons, 1974), p. 33. 25Ibid., pp. 32-33. 26Ibid., pp. 33—35. 27Everett Rogers and F. Floyd Shoemaker, Communi— cation of Innovations (New York: The Free Press, 1971), pp. 350-51. 28Bradley S. Greenberg and Hideya Kumata, "National Sample Predictors of Mass Media Use," Journalism Quarterly 45 ( ): 644—46. 29J. M. Nilles, F. R. Carlson, Jr., P. Gray, and 3. J. Hanneman, The Telecommunications—Transportation Fradeoff: Options for Tomorrow (New York: John Wiley & Sons, 1976), pp. 124—25. 3OIbid., p. 125. planni: fields freque: prior 1 They ge niques: l. A- Szal hein, K tiVeneSE actiVit§ aided de regard I DEhaviOI CHAPTER I I I METHODOLOGY The Research Technique ____________________ The research technique was borrowed from urban inning/transportation research. Researchers in these alds are primarily concerned with temporal (duration, :quency and timing of an activity) and spatial (location or to, during, and following an activity) variables. y generally utilize one of three data collection tech- ues: 1. Standard phone, mail or personal survey interview- ing techniques using questions of the form, "have you engaged in activity X and if so, from when to when or how often." 2. Self reporting of activities within a diary cover- ing all activities over a series of days. 3. Personal interviews using aided or unaided day— after recall of a single days' activities. lzalai, in collaboration with P. E. Converse, P. Feld— l, K. K. Scheuch, and P. J. Stone, analyzed the effec— ness of these three techniques in studies of human Vity patterns from twelve countries. They found d day-after recall to be the most successful in rd to most kinds of behavior.2 The one type of vior researchers had difficulty obtaining in this 82 fashic which front show u with 5 survey can be A time‘ this 8‘ detail( One ' lir§Ei£ Study V 83 shion was secondary activities (such as conversations) ich accompanied primary activities (like sitting on a ant porch or shopping). Secondary activities did not 3w up very well in diaries either.3 Since the present study was directly concerned :h such secondary activities, a standard telephone rvey using "how often" questions was conducted. Development of the final survey instrument (which 1 be seen in Appendix A) took place in four stages: 1. Selection of potential independent variable questions 2. Pretesting of potential independent variable questions 3. Pretest analysis leading to construction of the final instrument 4. Development of the dependent variable measures and placement in the final instrument imetable outlining the methodological activities of 5 study can be found on the next page. A more ailed discussion of these activities starts below. Selection of Potential Independent Variable Questions :tion Design The selection of potential questions for this .y was based upon the assumption that if each of the Weeks Week 5 Week 6 Week 7 Week 8 Week 9 Week 1 Wer 1 84 TABLE 4 TIMETABLE OF THE METHODOLOGICAL ACTIVITIES Time Activity :ks l — 4 Selection of potential independent variable questions 2k Pretest samples drawn Interviewer training for shopping pretest 2k Shopping pretest data collection including verification calls Interviewer training for banking pretest 2k Banking pretest data collection including verification calls Coding of shopping pretest data Keypunching of shopping pretest data begun Development of dependent variable measures Coding of banking pretest data Keypunching of banking pretest data begun 10 11 Pretest data analysis and construction of final questionnaire Discussion of revised instrument with interviewers Final sample drawn Final data collection Verification calls made Coding of final data Keypunching of final data begun Data processing begun nondem techno shOppi a wide ables : reflec unbala format and Me additi' To dev. initia cussed these shOppi; condUC‘ Michig. 0f nin: 0f the and of: 398810] Were t] CGIVed Oriente t0 ViE\ 85 demographic independent variables (attitudes towards hnology, attitudes towards shopping transactions, and pping and banking transaction behaviors) consists of ide range of attitudes or behaviors, then these vari— es must be measured via a series of questions adequately lecting that range. Choosing to adopt the primarily alanced, forced choice rating or Likert scale question mat used previously by Stone, Rich, Pace and LaPorte Metlay;4 an attempt was made to construct a five—point itive scale for each nondemographic independent variable. develop these scales, a large number of questions was tially drawn from the secondary source materials dis— sed in the first two chapters. The appropriateness of se secondary source questions in covering current day pping and banking transaction behaviors was checked by ducting two focus groups. One group contained nine higan State University students, the other consisted nine nonstudent adults. All subjects were residents the East Lansing area. They were recruited by phone offered a $5.00 incentive. The hour and a half sions were taped and transcribed. The transcriptions 2 then content analyzed. The findings confirmed the idea that people per— Ved differences in impersonal versus interpersonal :nted transactions. For example, some subjects seemed 'iew going to a store or bank as a necessary ev11. Shopp. neede< talk 1 as ea: usual] as fas Other time i 98t to things depres; conver: group. reporte cal bar range c that of t0 repl Dem .._£§pg§ ment Sta Order tc questioI attempt each den 86 pping trips preferably involved finding everything they ded quickly in one store. They expressed no desire to k to sales Clerks, and perceived phone or mail orders easy and convenient. Mechanical bank tellers were ally well received by these subjects who described them fast, simple to use, easily available, and convenient. er subjects, however, stated that they enjoyed spending e in stores; some of them using shopping as a chance to together with friends, or to take their mind off other ngs, "I always buy myself some new clothes when I'm ressed." A desire to shop in stores where they could verse with sales clerks was expressed by members of this up. Infrequent use of phone and mail orders was orted by these subjects and they tended to view mechani- 1bank tellers as impersonal and complex. Since the e of statements made by group members often resembled of the secondary source questions, it appeared valid :eplicate those questions in this study. >graphic Questions Last in the independent variable question develop— stage, the demographic questions were designed. In r to be consistent with the other independent variable tions (five—point scales), it was assumed that an npt would be made to collapse the response set of demographic question into five categories by approx could were d: categOJ Bureau knowlec status some cc to the less th studies for use referrec groups, to be us Variable identify Subjects and furt or fOur ThESe we. 87 tpproximating quintile splits. When no such divisions :ould be made between groups of responses, dummy variables were developed. The questions generally used standard response :ategories. That is, they were similar to U.S. Census Bureau formats, with slight adaptations made according to (nowledge of the population. For example, educational status classifications were reduced to college grad, some college, high school grad, and some high school due :0 the virtually complete absence of people possessing Less than some high school education found in previous studies conducted in the East Lansing area.5 In other cases, questions were developed especially for use in this study. For example, three questions referred to the presence of children within certain age roups, in a subject's household. These questions were 0 be used in the construction of the lifecycle status ariable. Another series of questions was developed to entify light to heavy travellers according to how often bjects reported traveling distances within five miles d further than five, ten, thirty, fifty, one hundred, four hundred miles from home during the previous week. ese were to be used in the calculation of physical bility variable scores. Prete: Varia] two uI Univex class shoppi graphi pated thorou proced interv Speaki viewer the ag. Verifi: back pg samPles because Surrour banks c that, b automat use of Variabl 88 Pretesting the Independent Variable Questions Pretest interviews were conducted by students in two undergraduate advertising classes at Michigan State University. In order to shorten the pretest process, one class used a questionnaire containing demographic and shopping items, while the other class worked with demo- graphic and banking items. All of these students partici- pated in interviewer training sessions, consisting of a thorough discussion of the questionnaire and interviewing procedures. Interviewers were instructed to conduct all interviews between 9 A.M. and 9 P.M. While preferably speaking to the person who answered the phone, the inter— viewers were told to attempt to screen out persons under the age of eighteen. To insure the validity of interviews, erification procedures, involving a minimum of one call ack per interviewer by the researcher, were explained. The pretest samples consisted of systematic random amples of two hundred phone numbers drawn from the Lansing uburban City Directory. The City Directory was used ecause it allowed easy access to East Lansing and the urrounding townships, an area within which there are two anks offering automatic teller services. It was believed hat, beyond reaching a potentially greater number of utomatic bank teller users (which was desired due to the e of impersonal banking behavior as an independent riable), this population base would not bias the researl area p2 and/or forty-. PreteS‘ by the Sciencl of the demogr. to be . items ' form 0. Since relatL very 1 Propos. retain was th quenci rePort refuse “Snail th6y h HEgati 89 research findings any more than the use of the general area phone directory. After the rejection of incomplete and/or invalid interviews, final samples of fifty and forty-five were obtained. Pretest Analysis All of the pretest data were coded and keypunched by the researcher. The Statistical Package for the Social Sciences (SPSS) was used first to obtain a frequency count of the responses to all questions. A check on the pretest demographic question distributions showed the two samples to be quite similar in composition. Next, all scaled items were paired off and zero order correlations in the form of Pearson Product Moment Correlations were computed. Since 97 percent of the interscale and intrascale cor- relations were below the .2 level, there appeared to be very little, if any, dependence between and across the proposed scale items. Therefore, these items were retained for use in the final questionnaire. The only problem encountered at the pretest stage has that interviewers reported, and the missing data fre— Iuencies confirmed, a high percentage of refusals on self- reporting banking behavior questions. The most often ‘efused questions were those dealing with where subjects Lsually conducted their banking transactions, i.e., where hey held checking and/or savings accounts. To keep the egative feelings associated with these questions from harmi respo of th struc appro Sever via T read indiv broad used for T This the 5 cover amoun t0 pa dSSQm into Varia Lastl 90 rming the overall response rate or biasing other ;ponses, these questions were placed at the very end the final questionnaire. Development of the Dependent Variable Measures One of the most difficult activities in the con— :uction of the survey instrument was the development of >ropriate measures of favorability towards TPDS. reral Verbal descriptions of the services available 1 TPDS were written. These alternatives were then id to and discussed with an opportunity sample of ten lividuals, including students and nonstudents within a >ad age range. The comments of these individuals were ad to draft the version found in the final instrument. To assess potential TPDS usage while controlling ' TPDS value, a five—point scale question was designed. ,5 question asked how often subjects would use TPDS if ‘ service was available at no cost. The value issue was ered through a question which asked for the maximum unt, within a O to $50 range, subjects would be willing pay on a monthly basis for access to TPDS. It was umed that these dollar responses could be collapsed o a five-point scale and tested against the independent iables in the same manner as the usage responses. :ly, to serve as a descriptive measure, a question COHCGI have n EEEJE Lansil frame ing a] temat Ibe_£ Michi the f of th needs To st were be 0; by th Calls Data Studx Conic SPEcj 91 )ncerning the type of product one would most want to ive made available via TPDS was designed. Final Data Collection 1e Sample Following the precedent set by the pretests, the ansing Suburban City Directory was used as the sampling rame. Allowances were made for a larger sample by draw- ng approximately seven hundred phone numbers on a sys- ematic random basis. he Interviewing Process Students in an advertising research class at ichigan State University were called upon to conduct he final interviews. Having already been briefed in all f the standard telephone interviewing procedures, they eeded only an introduction to the revised questionnaire. 0 stress the importance of valid interviews, students ere reminded of the verification procedures they would 2 operating under. A total of 673 phone calls was made { the interviewers. The final distribution of these ills can be seen in Table 5. ita Analysis The major objective of data analysis in this :udy was to determine whether favorability towards TPDS luld be expected to vary among individuals according to ecific demographic attitudinal and behavioral charac the re sis, v Cess l items Prese: jects Willi cerni GiVen of hy all f prESe 92 racteristics. In order to make this determination, research data went through two stages of data analy— , variable construction and hypothesis testing. TABLE 5 DISTRIBUTION OF SAMPLE PHONE CALLS Outcome Frequency Percentage Disconnected 30 4 Busy 3l 5 No Answer 253 38 Refused 61 9 Incomplete l9 3 Completed 279 41 Total 673 100 ‘iable Construction The first step of the variable construction pro— :s was to obtain frequency runs on the questionnaire ms. Looking initially at the number of missing cases sent for each item, it was found that eighteen sub— ts failed to respond to either the question regarding lingness to use TPDS "for free" or the question con- ning how much one would be willing to pay for TPDS. en the importance of responses to these items in terms hypotheses testing, these subjects were deleted from further analysis, leaving a final sample of 261. After rerunning the frequency distributions for remaining sample members, the number of missing cases :ent in all of the independent variable items were examix side 1 shOppi two mi missir T0 kee they V freque were: 93 nined. Only the items related to hours of work out— 2 the home by women, ages of children, impersonal pping and banking, and travel mileage had more than missing cases. These items were expected to have sing cases because: 1. Only females responded to the hours worked question 2. Children were not present in all sample house- holds, although less children were present than had been expected 3. Impersonal banking transactions are not available at all banks 4. The shopping questions were asked only of those people who reported having shopped by phone or mail in the last three months 5. The travel questions were sequenced so that once the subject's maximum travel level was reached, the remainder of the travel questions were skipped ceep these items from biasing further data analysis, 1 were recoded in their proper variable forms, and [uencies were recalculated. The coding changes made distri} shOppi] distri] shOppi! tribut: PortiOI 94 l. The working women item was made into a dummy variable. Thus all subjects were coded as being (work more than ten hours per week) or not being working women 2. The lifecycle status variable was made into a dummy variable by combining marital status and the presence or absence of children in the house- hold rather than age of children 3. Given the filtering used in the questionnaire in regard to the travel mileage, impersonal shopping and banking questions, it was assumed that the most appropriate classification for the missing cases was "never use." The five—point scale as opposed to conversion to a dummy variable was used in order to save as much information as possible Looking next at the shape of the item response :ributions, all items but those dealing with impersonal >ping and banking were found to have reasonably normal :ributions (mean = median = mode). The impersonal Iping and banking items had negatively skewed dis- utions, meaning that, not unexpectedly, a large ion of the sample reported "never using" impersonal saction modes. separ varia ficie were « or in' ducte< a guic ships. the tv Most c to be some 5 but 51 0f the Coline Counte Demo re V ' . we; tranSfc Pendent COnSist SOCiQeC represe 95 The second step of variable construction was to aparate all questionnaire items into their respective :riable groups. Pearson product moment correlation coef— Lcients, with pairwise depletion and one-tailed t—tests are calculated between all items to check for interitem : intrascale colinearity. One—tailed tests were con— lcted because the research hypotheses could be used as guide for predicting the direction of item relation— lips. The strongest correlation found was a .41 between 1e two dependent variable measures TPDS usage and value. >st of the remaining correlation coefficients were found 3 be below .25 in value (see Appendix B). There were Dme significant correlations above .25 within scales, it since these were always between multiple measures 5 the same scale component, the presence of interitem >linearity, as.a source of statistical bias, was dis— >unted.6 Variable Composition mographic Independent riables Various data manipulations were made in order to ansform the questionnaire items into the desired inde- ndent variable measures. The socioeconomic status data, Isisting of occupations which were assigned Duncan :ioeconomic Index scores,7 were divided into quintiles >resenting higher status to lower status. The educa1 when i were c were c Questi and th in the The fo their 1 were: dren, Childre from yc item sc for eac time su eXposur number Variabl. Cable t, Was alsc trewellE Calculat Subject: PHY 96 educational level responses broke quite evenly into thirds when the same high school and high school grad categories were combined. Thus, the educational status levels used were college grads, some college, and any high school. Questions regarding the marital status of respondents and the ages of children in their households were combined in the construction of a dummy lifecycle status variable. The four categories of this variable in rank order of their hypothesized relationship with TPDS favorability were: (1) single with children, (2) married with chil— iren, (3) single without children, and (4) married without :hildren. The age data were split into quintiles ranging from youngest to oldest. All social participation scale .tem scores were added together to obtain a single score 'or each subject. Totals were computed for the amount of .ime subjects reported using TV and radio (broadcast Xposure) and for the time spent with newspapers and umber of magazines read (print exposure). A dummy ariable representing present or past experience with able television versus no experience with cable television is also developed. The questions dealing with distances :avelled by subjects during the last week were used to llculate a single physical mobility score for each lbject: PHYSICAL MOBILITYi = (Ai*2) + (Bi*5) + (Ci*lO) + (Di*30) + (Ei*50) + (Fi*100) + (Gi*400) where The u ing f lowes 97 where: A - G = the parts of question #52 of the question— naire contained in Appendix A. i = values corresponding to an individual subject. The use of this formula resulted in a set of scores rang— ing from the highest mobility status level down to the lowest mobility status level. Behavioral Independent Variables The responses to the sets of questions regarding how often subjects conducted impersonal banking trans— actions, interpersonal banking transactions, and impersonal shopping transactions were totalled, providing one score ‘for each type of transaction per subject. No manipulation was needed in regard to the item measuring how often sub- jects conducted interpersonal shopping transactions or those items dealing with when such transactions usually occurred. Attitudinal Independent Variables The four items in the attitudes towards technology scale were added together to obtain a single score for each subject. The attitudes towards shopping questions (shopping orientation scale) required the greatest amount of me was p six f facto essen loadi as th other aPaths going Chase amOunt meanin 98 f manipulation. Principal factoring without interaction as performed on these items. The items broke down into ix factors explaining 63 percent of the variance. The actor matrix can be seen in Table 6. In most cases ssentially one or two items dominated (have factor oadings > .5) each factor. The first factor while having the browsing item 3 the heaviest loading, also had high loadings on six ther items. These items were interpreted as reflecting: 1. Positive attitudes towards shopping as a social activity involving interpersonal communication. An opportunity for companionship and conversation with others 2. Positive attitudes towards product pricing activity, meaning comparison shopping and bargain hunting 3. Negative attitudes towards shopping on a time or travel efficient basis The second factor indicated a high convenience or athetic orientation; that is, positive attitudes towards ing to the closest store that allows one to quickly pur— ase everything one needs. Yet there was also a certain ount of price consciousness. This was interpreted as aning that the store chosen as being highly convenient w mHmANH m.~e emansssooa m.s m.k m.m o.a s.NH m.wH wasHmHaxm mosmflum> mo mmoucoouwm 0N.: 0H. mo.u me.- mmm «H.u mm. moaaumHmmm meHouumH mH.- mo.- so.u mo.n .mmw o .n we. muonm ummmoHouuwH no. No.1 em. om. bub km.u we. norm donmuoaolnaH em. mH. eo.- VN. Hm. mv.n we. msHu mmsHm>-an om. Ho. mo.» as. HH.- mmw ma. nHms n.2mou|mH HH. aH.- mo.- mN.- mm. as. mm. Amxano nose mudgu0\3 mmno>soollva Hm.u MMh mm. sm.- mo. ow. am. m>HmHomeaH|INH % «H. am. Mbh- Ho.- No. om. Hm. xan0\3 wmnm>coo|IOH me.- oo.- mp. MHh mo. pm. as. w>HmHsasHuum mm.n mH. me. am. .pww Hmw an. nso semmHH<--m mo. mN.u NH.- so. mm. new mm. mcanse sHmmnmmula kH.- km.- 00. OH. No. mph em. qumzonmlum mwhu wH. mm.u NH. em. emu we. monm somHnaasooulm am. «0.- NH. mm.: mm. as. so. mnwan0\3 donmlum m m a , m N H muflamcsfifiov vaH mnflocsoflpmmoo mHOpowm paw mmcflpmoq Houomm QWZHfiAmxm HUZ¢Hm¢> mo BZDOE< 02¢ mOZHmfloq MOBUrm, Pearson Product Moment Correlations were computed atween all independent variables to check for multi— >linearity. The coefficients can be seen in Appendix B. rer 95 percent of the correlations were below the .l avel with the highest being a .34 between educational 1d socioeconomic status. Given that Kim and Kohout Late that extreme multicolinearity is reflected by cor- lO lations 3 .8, it was believed that the data were latively free of this source of error. Eothesis Testing As stated in Chapter II, the research hypotheses ced the independent and dependent variables within ee levels of analysis. The first level consisted of ariate relationships, with each dependent variable ng paired off with one independent variable at a time. pin's model suggests that such relationships will be weak. That is, interaction between two or more inde} to e; opera expec in a of ve engag Here, in ei nents pende varia to en large Throu Of th expec each. the II atiOn regre. with a 103 independent variables appears to contribute more variation to either dependent variable than any independent variable operating alone. Therefore, bivariate relationships were expected to explain only a minimal amount of the variance in a dependent variable. At the second level, clusters of variables (propensity to engage and opportunity to engage) were tested against the dependent variables. Here, each cluster was expected to explain more variance in either dependent variable than its individual compo- nents. The third level involved testing all of the inde— pendent variables simultaneously against each independent variable. This joint action of propensity and opportunity to engage was expected to contribute to or explain the largest amount of variance in either dependent variable. Throughout all of these levels of analysis, the direction of the relationships, as specified by the hypotheses, was expected to hold true. Bivariate or multiple regressions were run for each pair or group of independent variables presented in the research hypotheses. In those cases where consider- ation was given to more than one independent variable, the standard regression method was used because no pre- etermined basis existed for adding variables into the . . ll egreSSion equation. Hypothesis testing at the bivariate level started ith an examination of the significance levels attributed to t) BiseJ lines The s and c while estin depez varia anal) hypot due c When cont: regre hypot the i eQuat relat Beta regre measu beta Varia 104 to the simple r's (Pearson Product Moment or Point- Biserial Correlations) and F—tests which measured the linear associations present in each regression equation. The simple r values were then used to assess the strength and direction of all significant relationships (p i .05); while the coefficients of determination (r2) provided estimates of the amount of variance explained in the dependent variable.12 Hypothesis testing of the multiple independent variable level hypotheses involved several stages of analysis. First, an F—test was used to test the null hypothesis that any observed multiple correlation was due only to sampling fluctuation or measurement error. When this null hypothesis was rejected (p i .05), the contributions of the independent variables to the multiple regression equation were assessed to see whether the hypothesized relationships were upheld. In order to determine the actual contributions of the independent variables to the multiple regression equations, the beta weights and squared semipartial cor— relations of the independent variables were examined. Beta weights were used, as opposed to the unstandardized regression coefficients, because the independent variable measurement units were not always of the same form. The beta weights indicated how much change in the dependent variable was produced by a standardized change in a cert othe Assu dent size wroug part: deper pends expre ment (tota deriv all 0- Sidere the if Cantly for a t0 be Presen 105 certain direction by an independent variable when all other independent variables were being controlled. Assuming that no colinearity existed among the indepen- dent variables, this meant that the greater the absolute size of the beta weight, the greater the amount of change wrought in the dependent variable.13 The squared semi- partial correlations were used to measure how much of the dependent variable variance was explained by each inde— pendent variable. The squared semipartial correlations expressed this information in terms of the absolute incre- ment of the squared multiple correlation coefficient (total amount of variance explained by regression equation) derived by the addition of an independent variable when Iall other independent variables had already been con- sidered.14 Given a significant multiple correlation, all of the independent variables had to be contributing signifi- cantly, in the proper direction, to the regression equation for a multiple independent variable research hypothesis o be fully accepted. The results of all of these hypothesis tests are resented in Chapter IV. CHAPTER I I I --NOTES lAlexander Szalai, ed., The Use of Time-Daily Activities of Urban and Suburban Populations in Twelve Countries (The Hague: Mouton and Co., 1972), p. 74. 21bid. 3F. Stuart Chapin, Jr., Human Activity Patterns in the City: Things People Do in Time and Space (New York: John Wiley and Sons, Inc., 1974), p. 192. 4G. Stone, "City Shoppers and Urban Identification: Observations on the Social Psychology of City Life," The American Journal of Sociology 60 (July 1954): 37-45; Stuart U. Rich, Shopping Behavior of Department Store Customers (Boston: Harvard University Graduate School of Business 'Administration, 1963); C. R. Pace, They Went to College (Minneapolis: University of Minnesota Press, 1941); Todd R. LaPorte and Daniel Metlay, "Technology Observed: Atti- tudes of a Wary Public," Science (April 17, 1975): 121—27. 5"Faygo Softdrinks, A Consumption Study for Primary and Secondary Markets" (unpublished report made to the Faygo Company by Martin P. Block and Maureen E. Beninson). 6Jae-On Kim and Frank J. Kohout, "Multiple Regres- sion Analysis: Subprogram Regression," in Statistical Package for the Social Sciences, eds. N. J. Nie, C. H. Hull, J. G. Jenkins, K. Steinbrenner, and D. H. Brent (New York: McGraw Hill, 1975), p. 340. 7John P. Robinson, Robert Athanasiou, and Kendra . Head, Measures of Occupational Attitudes and Occu- ational Characteristics (Ann Arbor: Survey Research enter, Institute for Social Research, 1974), pp. 343-56. 106 107 8Jae—On Kim, "Factor Analysis," in SPSS, pp. 488-89. 91bid. 10Kim and Kohout, "Multiple Regression," p. 340. llIbid., p. 336. 12Ibid., pp. 330—31. l3Ibid., p. 332. 14Ibid., p. 333. CHAPTER IV FINDINGS Before entering into a discussion of the hypothesis test results, the sample demographics, descriptive question responses and the frequency distribution of the dependent variable measures will be presented. As discussed in Chapter III, this study's sample was drawn from the Lansing Suburban City Telephone Directory. The sample, while con— taining some biases in regard to making nationwide gen— eralizations, was assumed to be representative of the population area. That is, while some differences might exist between people whose phone numbers were present or absent from this sampling frame, since the number of excluded households was reasonably small (20 percent), a random sample drawn from this frame could be expected to oe demographically similar to the entire population of the East Lansing area.1 The details of the sample's impersonal shopping transactions were examined in an attempt to incover any behavioral trends among heavy vs. light users. he dependent variable measures are presented to identify he range of the sample's reactions towards the TPDS concept. 108 isti' samp; spli- sampl and 1 about of t} and t there educa sampl Ratin was b biase Michi Effec follo bilit SOnal betwee 109 The Sample The distributions of the demographic character- istics of the final sample can be viewed in Table 7. The sample consisted primarily of females, with a fairly even split between single and married people. Nearly half the sample lived in households containing two to three people and there were children under the age of eighteen in about one-fourth of the sample households. A majority of the sample was under thirty years of age (58 percent) and had at least some college education (79 percent), thereby biasing the sample towards younger, more highly educated individuals. Approximately three-fourths of the sample had a middle or upper Duncan Socioeconomic Status Rating. Therefore, the socioeconomic status distribution was biased in favor of the higher levels. None of the biases found were unexpected given the proximity of Michigan State University to the population area. The effect that these biases might have on the study findings, following the views of Katzman, would be a stronger favora- bility towards TPDS by the sample as a whole. Descriptive Questions--Sh0pping by Phone or Mail Initially frequency counts were run on all imper- sonal shopping transaction items. Then crosstabulations between the reported frequencies of mail or phone shopping Demogra] Sex Marital Educatic Age Socioesc Status Size of Age of c in House Clerks, . Rating 3 makers, , and depa: 110 TABLE 7 DEMOGRAPHIC CHARACTERISTICS OF THE SAMPLE Population Demographic Variable Classification Frequency Percentage Percentage (N=45357) Sex Female 180 69 51 Male 81 31 49 Marital Status- Single 136 52 56 Married 125 48 44 Education College grad 89 34 34 Some college 117 45 52 High school grad 42 16 10 Some high school 13 5 4 Age Under 18 yrs. 10 4 ~— 18 to 24 98 37 42 25 to 30 44 17 12 31 to 45 52 20 20 46 to 65 44 l7 19 Over 65 13 5 7 Socioeconomic Lower status 71 27 -- Status Middle Status 91 35 —- Upper status 99 38 -- Size of Household 1 person 28 11 ~- 2 to 3 people 117 45 -- 4 to 5 peeple 83 32 -- 6 or more people 33 13 -- Age of Children Under 13 yrs. 25 10 -- in Household 13 to 18 yrs. 23 9 —- 1 day to 18 yrs. 19 7 —- No children 194 75 -- Total 261 100 * Lower status = Duncan Rating < 33 (includes porters, shipping lerks, gas station attendants, machinists, etc.); Middle status = Duncan ting 34 to 66 (includes nonauto mechanics, secretaries, tool and die akers, salesmen, etc.); Upper status = Duncan Rating > 66 (includes buyers nd department heads in retail stores, teachers, doctors, etc.). and g Squar In or quenc fregue ping) questi totals chase the ta the ch are d1 was no Blalocj Was be. result Small 5 Provide lEVel C invOlve 111 and purchase characteristics were analyzed using Chi- Square with a missing values adjustment and Cramer's V. In order to reduce the number of empty cells, the fre- quency of usage items was collapsed into three categories: heavy user (very often + often) light user (not often + not very often) nonuser (never) The crosstabulation tables had very small total frequencies (N=63 for phone shopping, N=75 for mail shop— ping) because the nonusers did not answer the descriptive questions. Generally these tables had much larger marginal totals for the frequency of use categories than the pur— chase characteristic categories. The Cramer's V measure of association was used on the tables because this statistic makes an adjustment in the chi-square, correcting for the fact that its values are directly proportional to the number of cases.2 There was no manipulation of the cell data, as suggested by Blalock when marginal totals greatly differ, because it as believed that too much information distortion would esult from such manipulations.3 Given that with the mall sample size, only very strong relationships would rovide significant (p < .05) chi-square values,4 the evel of the Cramer's V was examined for all tables. In a few cases where all interval level data were nvolved, the strength of the relationship was measured by P! the c concj Ideni Mail or me phone relat table 14 pe mail Pho: Shoppe Heavy Light Non TOta 112 by Pearson Product Moment Correlations. A discussion of the descriptive question findings is provided at the conclusion of the data presentation section. Identification of Phone and Mail Shoppers About two out of every five (43 percent) people in the sample were found to have previously shopped by phone or mail. The relationship existing between frequency of phone and mail shopping (see Table 8) had a Pearson cor— .19 As shown in the relation coefficient of (p=.OOl). table, 19 percent of the sample shopped by mail only, 14 percent by phone only and 10 percent had used both mail and phone shopping services. TABLE 8 PEARSON CORRELATION: PHONE AND MAIL SHOPPERS Shggggis Heavy Light Non Shggpérs Heavy 4(2%) 5(2%) l6(6%) 25(10%) Light 3(1%) 14(5%) 21(8%) 38(15%) Non 8(3%) 41(16%) 149(57%) 198(76%) Totals 15(6%) 60(23%) 186(71%) 1(100%) NOTE: r = .19, p = .001 Sour ticu info Tabl tive refe; expo: of m1 indi< varie Phone menti by pt Catdll NeWSpg Magaz: Mail ( PriOr Telev: Other Tot 113 Sources of Information The users reported that print media forms, par- ticularly catalogs, were more common sources of product information than broadcast media or other people (see Table 9). In fact, radio was never mentioned. A rela— tively large number of people cited "prior exposure" as a source when shopping by phone. This category primarily referred to personal knowledge resulting from in-store exposure or previous use. The difference in the number of multiple responses relative to sample size seemed to indicate that people who shopped by mail used a greater variety of information sources than those shopping by phone. No relationships were found between the first mentioned information source and frequency of shopping by phone (see Tables 10 and 11). TABLE 9 ALL MENTIONED SOURCES OF INFORMATION FOR PRODUCTS PURCHASED BY PHONE OR MAIL Phone (n=63) Mail (n=75) #Mentions Rank #Mentions Rank Catalogs 33 l :3 2 Newspapers 25 2 22 2 Magazines 9 4 9 3 Mail Circulars 2 7 13 7 Prior Exposure 12 3 4 6 Television 5 5 9 5 Other People 3 6 Total 89 —— 123 -- NOTE: Spearman's Rho = .25. 114 .HHm. u >\mo. A a .Hem .NH.m n Nx "msoz HNOOHV HmeHe Ammo Heme Hesse Ammo Hmva me oH N m mN N mH Hamel Ammo HNme Ammo meNv Ammo AWNHV as e N m SH H HH nrmHH Hmkmv Ammo AsmHv AsHHV MN a o o NH o N s>mmm mousom wudmomxm wamowm COHmH>m m moo m. m mosHmnmm mHmQMQmb mummmoam coflmeHOmHsH HOHHm Hmcuo . . H B H p U . Z Z msogm mzomm em ozHaaomm ozd momDOm oneazmomzH omonszms smmHm "mm\mo. A a .ues .NH.N u Nx "meoz ANOOHV Haas Haas Ammo HWGHV ANNav HmmHv HmeHv mm m m N NH Nm HH NH menv Ammv vav mev ANNHV meNv ANNHV Amev mm N m N a NN a NH namHH ANHNV ANHV Awev ANMHV mev 0H H o o m 0H N o N>mwm mossom whomomxm mHmowm sOHmH> mHMHsoHHU moo m m mmsHm mucosa mnwmmocm COHHoEHomsH HOHHm Hmnuo IonB Han H p 0 Icon: Imswz HHmz ”mmmDOmIHmo HHfiE Wm OZHmmomm QZd MUMDOm ZOHBdSMOhZH DMZOHHZME BmmHm HH mqm\mo. A a .uee .mm.m n Nx "msoz ANOOHV mev Away ANMHV vaHv wav Awmv Hwowv mm m m m m m m mN Awmmv $3 $3 $3 $3 $03 $3 $.va 04 m a m m m m mH urmHH Awnmv mev Awmv Awwv mev Ammv meHV mN o N m a N N OH N>awm llllllllllllll munnm pHoz mcsoomm w mHOOB uWthu Immsom hnwoosw waHNmmnz sconce mmpooom CMHMBH msHfluoHo mwmmmonm p m Heumcow mxoom Ugo Huuo m I a u.M Sm pHonmmsom mzomm Nm OZHmmomm Q24 WAWB Ebbmomm "mm\mo. A a .meo .ma.H u Nx "msoz me e m mH n VH w eN menv vav mev Hwomv mev ANMHV Hwev vamv mm m N mH s OH m NH nrmHH HAHNV “NHV ANHV Awmv Hva Ammv Hva mev SH H H a H N H e N>mwm mgsom pHos mpuoowm w mHOOB wmhe Iwmsom hsmoosw mwsflummmz smpsnw mpoow mums mcHnuoHo mchQonm pooponm Hohwcmw mxoom one mCHHHomm Ismaopflm . HHmE UHoamwsom HHmz “WWZ > cmuwo uoz Gmumo MHWWO mconmz awe and Sea mzoma em noun 2. mzaHa masses "onsaHmmmoo zommamm mH mqmae 121 SOOHO $2 Home THE SOHO $3 E N Na OH OH m 3323 $ch $233. ANNHONH $.33. O 6.:qu HAHNOOH O SNON Sim HBO seem Same socwwwmmfiwmwa .2... 3%ng is .02 is “we .Mmmmma mMD Bmflm 02d HH<2 Mm momm OB mZ.05, rv2=.00). Furthermore, the small association that was found was not in the direction that had been predicted meaning that favorable attitudes towards technology had a slight ten— dency to be related to attitudes that viewed TPDS as low in value. Together these findings may mean that while people who are generally favorable towards technology would like to use TPDS, they would not necessarily be willing to pay a lot for such a service. Hypothesis 2: People having more positive attitudes towards imper- sonal shopping transactions will be more favorable towards TPDS than those having more negative attitudes. This hypothesis was not supported. Impersonal shopping attitudes, as discussed in the previous chapter, were represented by three of the shopping orientation scale were value fact in t mean comm perc aspe less 128 scale factors (factors 2, 3 and 4). None of these factors were found to have significant relationships with usage or value (see Table 20). Additionally, the sign of the factor 3 usage coefficient portrayed a relationship going in the opposite direction of that hypothesized. This may mean that while some people seek to avoid interpersonal communication oriented shopping behavior, TPDS is not perceived as a viable means of avoidance. That is certain aspects of TPDS may have raised conflicts resulting in less favorability than expected. TABLE 20 PEARSON CORRELATION COEFFICIENTS FOR IMPERSONAL SHOPPING FACTORS Factor 2 (high convenience, price conscious orientation) .08 >.05 .01 .02 >.05 .00 Factor 3 (negative towards clerk assistance, tends to shop while already Out) —.01 >.05 .00 .07 >.05 .01 Factor 4 (impulse shopping) .003 >.05 .00 .04 >.05 .002 People scoring highly on factor 2, which did have the strongest correlation with TPDS usage, may have per— ceived TPDS as a high cost means of shopping. People sco: too the ind Peo as act bee tov of um co- re CO 129 scoring highly on factor 3 could be perceiving TPDS as too in-the-home oriented to get much use out of it. Yet there is a slight sense of value expressed, perhaps indicative of an appreciation of salesclerk absence. People scoring highly on factor 4 may not perceive TPDS as an environment permitting much impulse shopping activity. Such negative perceptions of TPDS could have been strong enough to outweigh any favorability produced towards TPDS by its impersonal communication orientation. Viewing TPDS favorability on a more long-term basis, some of these negative perceptions might be altered if people understood TPDS more fully. For example, service costs could be presented in terms of their cost efficiency relative to in—store shopping, thereby reducing the high cost problem. Hypothesis 3: People having more negative attitudes towards inter— personal shopping transactions will be more favorable towards TPDS than those having more positive attitudes. This hypothesis was partially supported. Examining the three remaining shopping orientations discussed in the previous chapter, two significant relationships were found (see Table 21). One of these relationships, between factor 1 and usage, did not, however, reflect the pre— dicted direction. The relationship found seems to indi- cate that TPDS was perceived as providing an acceptable 130 means of browsing, comparison shopping, or bargain hunt- ing. This form of favorability may have been either strong enough to overcome conflicts regarding social interaction, or was complemented by the idea of substi— tuting social interaction in-the—home for that occurring in stores. The lack of concern over time and travel might not have affected the situation at all. Yet favorability, in general, was not so great that TPDS was Viewed as being significantly more valuable by people scoring highly on factor 1 than those having lower scores. TABLE 21 PEARSON CORRELATION COEFFICIENTS FOR INTERPERSONAL SHOPPING FACTORS 2 2 ru Pu ru rv Pv rv Factor 1 (strong browser, price conscious, not time or travel conscious, involves social interaction) .15 <.05 .02 .02 >.05 .00 Factor 5 (discuss nonshopping topics with salesclerk) -.12 <.05 .Ol -.04 >.05 .002 Factor 6 (shopping is a way of getting out of the house involves social interaction) —.06 <.05 .004 -.08 >.05 .01 The second significant relationship found was between TPDS usage and factor 5. People having low 131 factor 5 scores (negative attitudes) were those who reported having no or much more infrequent conversations with salesclerks about nonshopping related topics. If people desiring such conversations (positive attitudes) are less willing to use TPDS, the idea that shopping ful— fills certain social or interpersonal communication needs is confirmed. Meaning that there are, indeed, some dis- tinctions between the motivations behind shopping and buying behavior. The correlation between TPDS value and factor 5 also went in the predicted direction but was not significant. Thus, there was only a slight tendency for people, not seeking nonshopping related conversatiOns with salesclerks to value TPDS more than people seeking such conversations. Low scores on factor 6 (negative attitudes) were received by people who tended not to view shopping as pri— marily a means of getting out of the house and enjoying the company of others. As predicted, this attitudinal position was negatively related to TPDS favorability in terms of both usage and value, but not significantly. Thus, there was only a slight tendency for people who view shopping as a means of participating in social inter— action to reject TPDS as a Viable form of shopping or buying. 132 Summary of Attitudinal Bivariate Level Hypothesis Testing In summary, people having positive attitudes towards technology may be expected to be more favorable towards the idea of using TPDS, but may not be willing to pay a lot for TPDS access. The issue of avoiding or having the opportunity to engage in interpersonal commun— ication does appear to have some impact on TPDS usage, but not as much as had been expected. The same issue did not seem to affect perceptions of TPDS value at all. It appears as though other characteristics of shopping or buying behavior may be fairly important. People did per— ceive TPDS as a viable means of browsing or comparing product attributes and prices. It is uncertain whether they would be willing to substitute shopping trips or in—store social interaction with others for interaction with others while shopping Via TPDS. It is not difficult to understand why people desiring nonshopping related social interaction with salesclerks were opposed to using TPDS since TPDS would not supply such communication. These people would pro- bably have to find, if possible, an acceptable alternative source of interpersonal communication before becoming favorable towards TPDS. Hypothesis 4: Women will be more favorable towards TPDS than men. 133 This hypothesis was not supported. Women did not wish to use TPDS significantly more frequently than men 2 . (ru=.01, pu>.05, ru =.00). Nor did women place signifi— cantly more value upon TPDS than men (rv=.02, pV>.05, rV2= .00). Hypothesis 5: People having a higher socioeconomic status will be more favorable towards TPDS than those having a lower SOCioeconomic status. This hypothesis was not supported. People of a higher socioeconomic status did not significantly express a desire to use TPDS more often than members of lower socioeconomic status groups (ru=.03, p>.05, ru2=.001). Nor was any significant difference found between socio— economic groups in terms of perceptions of TPDS value 2 (rv=.02, p>.05, rv =.00). Hypothesis 6: Working women will be more favorable towards TPDS than women who do not work. This hypothesis was not supported. Working women did not wish to use TPDS significantly more frequently 2 than other sample members (ru=.02, p>.05, ru =.00). Nor were working women willing to pay significantly more for 2 . . . TPDS than others (rv=.004, pv>.05, rv =.00). This finding was unexpected given the great travel and time advantages 134 TPDS offers working women. It is possible that the low number of working women in the total sample (64 out of 261) biased the correlation statistics. Hypothesis 7: People having a higher educational status will be more favorable towards TPDS than those having a lower educational status. This hypothesis was partially supported. People having a higher level of educational status did not wish to use TPDS significantly more often than those of lower educational status (ru=.01, pu>.05, ru2=.00). However, people of a higher educational status did perceive TPDS as being significantly more valuable than others of a lower educational status (rv=.l7, pv<.05, rV2=.028). Hypothesis 8: People having an early lifecycle status will be more favorable towards TPDS than those having a later lifecycle status. This hypothesis was partially supported. People having an earlier lifecycle status did not wish to use TPDS significantly more frequently than people of a later lifecycle status (ru=.06, pu>.05, ru2=.004). However, people of an earlier lifecycle status did perceive TPDS as being significantly more valuable than people of a 2 later lifecycle status (rV=.20, pv<.01, rV =.04). 135 Hypothesis 9: Younger people will be more favorable towards TPDS than older people. This hypothesis was supported. Younger people did express a desire to use TPDS significantly more often than older people (ru=.ll, pu<.05, ru2=.01). Younger people also perceived TPDS as significantly more valuable than . 2 did older people (rv=.26, pv<.01, rv =.07). Summary of Test Results for Hypotheses 4—9, Demographic Characteristics Reviewing the findings of the last six hypotheses, only one predictor of TPDS usage, age, was found. That is, younger people were found to be more favorable towards TPDS than older people. Three predictors, age, lifecycle status and educational status, were found to be related to TPDS value meaning younger, more highly educated people who are single and/or have children under the age of eighteen could be expected to be willing to pay more for TPDS than others. While none of the other correlations were significant, they all portrayed relationships in the predicted directions. Thus, there was at least a slight tendency for working women of a higher educational and socioeconomic status and a lower lifecycle status to be more willing to use TPDS than others; and for working women of a higher socioeconomic status to perceive TPDS as more valuable than others. 136 Looking back at the significant findings, probably the greatest discrepancy between them and the findings of others presented in Chapter II is the lack of influence of socioeconomic status. While many of the other variables varied from study to study, a higher socioeconomic status was nearly unanimously found to be related to favorability towards TPDS characteristics. It may be that this variable was not measured appropriately given the presence of some college students in the sample. At the same time, it may just not be as important a variable as it has in the past. Hypothesis 10: People having a higher social participation status will be more favorable towards TPDS than those having a lower level of social participation. This hypothesis was not supported. The correlation coefficients calculated in regard to social participation indicated nonsignificant relationships going in the Opposite direction from that predicted, meaning that people having a higher social participation status did not wish to use TPDS significantly less frequently than people of a lower social participation status (ru=—.06, pu>.05, ru2=.004). Nor did people having a higher social partic- ipation status perceive TPDS as significantly less valuable than people of a lower status (rV=-.08, pv>.05, rV2=.01). These hypothesis test results, although not sig— nificant, appear to suggest that people who are socially 137 active have a tendency to desire social interaction during shopping activities as well. At the same time people who are not socially active seem to have a tendency to prefer less social interaction in their shopping activities. Such trends would negate the possibility of people easily trading off the loss of social interaction in one situation for a higher level of social interaction in another situ- ation. Alternatively, these results may be reflecting differences in the amount of time people have available to participate in nonessential activities. In this case people who are less socially active may be operating under greater time constraints, and may View TPDS as being a more flexible, time—saving mode of shopping. Hypothesis ll: People having a higher level of broadcast media exposure will be more favorable towards TPDS than those having a lower level of broadcast media exposure. This hypothesis was not supported. The correlation coefficient calculated between broadcast media exposure and TPDS usage reflected a relationship going in the oppo- site direction of that predicted. However, people having a higher level of broadcast media exposure did not wish to use TPDS significantly less often than those having a 2 lower level of exposure (ru=—.Ol, pu>.05, ru =.00). People having a higher level of broadcast media exposure 138 also did not perceive TPDS to be significantly more valu- able than others having a lower level of exposure (rv=.03, 2_ pv>~05, rV —.001). Hypothesis 12: People having a higher level of print media exposure will be more favorable towards TPDS than those having a lower level of print media exposure. This hypothesis was not supported. Both of the correlation coefficients produced in regard to print media exposure suggested relationships going in the opposite direction from that hypothesized. Furthermore, people having a higher level of print media exposure did not wish to use TPDS significantly less frequently than people 2 having a lower level of exposure (ru=-.O3, pu>.05, ru - .001). Nor did people having a higher level of print media exposure perceive TPDS as being significantly less valuable than people having a lower level of print media 2- exposure (rV=-.02, pv>.05, rV —.00). Hypothesis 13: People having some experience with cable television will be more favorable towards TPDS than those haVing no experience with cable teleViSion. This hypothesis was not supported. People having some experience with cable television did not wish to use TPDS significantly more often than people lacking exper— 2 ience with cable television (ru=.05, pu>.05, ru =.003). 139 Nor did people having some experience with cable television perceive TPDS as being significantly more valuable than people lacking experience with cable television (rv=.06, 2 PV>-05, rv —.004). Summary of Test Results for Hypotheses 10-13, Media Use and Social Participation Looking across both print and broadcast media, people having less exposure to the mass media tended slightly to desire more frequent TPDS use than people having more media exposure. This type of relationship could, in support of the second explanation of the social participation test results, be indicative of people who, not having a lot of free time for social participation, or in this case, media exposure, would prefer to use TPDS because it could save them time or fit into their schedule better. The differences in the perceived value of TPDS relative to media use appear to highlight a possible trend in today's society in regard to media preference. Again none of the differences were found to be significant but there was a tendency for people having a high broadcast media consumption level and/or a low print media con— sumption level to value TPDS more highly. This could mean that certain people prefer broadcast media use and that TPDS, being broadcast media oriented, was thus more highly valued. However, since the number of people in 140 the sample who simultaneously fit into the high broadcast media and low print media consumption categories (r=-.19) was not that great, one cannot make generalizations about TPDS value and people who always favor broadcast media versus always favoring print media. Looking at the people who had experience with cable television, as one form of broadcast media con— sumption; these people did perceive TPDS to be slightly more valuable than those lacking cable television exper- ience. Thus, people having higher broadcast media con— sumption levels, particularly in terms of cable tele- vision, could be more favorable towards TPDS than others. Hypothesis 14: People having a lower physical mobility status will be more favorable towards TPDS than those having a higher physical mobility status. This hypothesis was not supported. Both of the correlation coefficients calculated reflected relation— ships going in the Opposite direction from that hypothe— sized. Furthermore, peOple of a lower physical mobility status did not wish to use TPDS significantly less often 2 than those of higher status (ru=.Ol, pu>.05, ru =.00). Nor did people of a lower physical mobility status per— ceive TPDS as being significantly less valuable than 2 . — =. . This means that others (rv—.05, pv>.05, rv 003) 141 people who travel a lot may be slightly more favorable towards TPDS than those who do not travel as much. Hypothesis 15: People having fewer cars in their household will be more favorable towards TPDS than those having more cars in their household. This hypothesis was not supported. In agreement with the physical mobility status variable, positive, rather than negative relationships were suggested by the correlation coefficients. However, people having less cars in their household did wish to use TPDS significantly 2 less often (ru=.10, pu=.05, ru =.01) and did perceive TPDS as significantly less valuable (rv=.l4, pv<.05, r 2:.02) than people having more cars. Summary of Test Results for Hypotheses l4 and 15, Physical Mobility The test results for Hypotheses l4 and 15 appear to indicate that people who are less mobile may wish to be more mobile; and may be negative towards TPDS since it is a less mobile activity, while people who are more mobile can perhaps see advantages to TPDS as a means of travel reduction. If a high physical mobility status, including access to a greater number of cars, can be equated with time restrictions, the above findings com— plement earlier bivariate analysis interpretations 142 regarding the possibility that people who have a limited amount of time available for shopping and other social activities may be more favorable towards TPDS. Hypothesis l6: PeOple who conduct more impersonal banking trans- actions will be more favorable towards TPDS than those who conduct fewer impersonal banking trans— actions. This hypothesis was partially supported. People conducting more impersonal banking transactions did not wish to use TPDS significantly more often than people con- ducting fewer impersonal banking transactions (ru=.01, p>.05, ru2=.00). However, individuals conducting more impersonal banking transactions did perceive TPDS as 2 more valuable than others (rv=.10, pv=.05, rV =.01). Hypothesis 17: People who conduct fewer interpersonal banking trans- actions will be more favorable towards TPDS than those who conduct more interpersonal banking trans- actions. This hypothesis was not supported. Although both correlation coefficients did express the predicted direction, neither was significant, meaning that people who conduct fewer interpersonal banking transactions wished to use TPDS slightly more often (ru=-.O6, pu>.05, 2 ru =.OO4), and perceived TPDS to be slightly more 143 2 valuable (rv=—.04, pV>.05, rv =.002) than those conducting more interpersonal banking transactions. Hypothesis l8: People who conduct more impersonal shOpping trans- actions will be more favorable towards TPDS than those who conduct fewer impersonal shopping trans— actions. This hypothesis was partially supported. PeOple who conduct more impersonal shopping transaction did wish to use TPDS significantly more often than people conduct— ing fewer impersonal shopping transactions (ru=.12, pu<.05, ru2=.01). However, people conducting more impersonal shopping transactions did not perceive TPDS to be significantly more valuable than others conducting such transactions less often (rv=,03, pv>.05, rV2=.001). Hypothesis l9: People who conduct fewer interpersonal shopping trans- actions will be more favorable towards TPDS than those who conduct more interpersonal shopping transactions. This hypothesis was partially supported. Four independent variables related to interpersonal shopping transactions were examined. The first variable, referring to how often one used a car when they went shopping, was found to have a significant relationship with usage (ru=.l7, pu<.01, ru2=.03), but this relationship went in the opposite direction from that hypothesized meaning that in support of the physical mobility status 144 conclusions, people who travelled to stores frequently wished to use TPDS significantly more often than those who used a car less often. The value relationship went in the predicted direction but was not significant (rv=-.04, pv>.05, rV2=.002). Thus, people who travelled to stores frequently perceived TPDS to be slightly less valuable than people who did not travel to stores as frequently. The second variable was concerned with frequency of shopping at night. People who shopped frequently at night did wish to use TPDS significantly more often than those shOpping less frequently at night (ru=.l7, pu<.01, 2 ru =.03). But frequent night shoppers did not perceive TPDS to be significantly more valuable than others 2 (rv=.07, pV>.05, rv =.01). It seems contradictory that more frequent night shOppers would express such a strong desire to use TPDS without a significantly higher per— ception of TPDS value. The findings may be reflecting the feeling that while TPDS offers certain advantages to people who fre— quently shop at night, the service is not so desirable that it is worth a lot of money. One rationale for a lower value perception would be that people are not ready to fully substitute TPDS for more traditional shopping modes. Instead, people View TPDS as an expensive alterna- tive in certain situations. 145 The third variable measured shopping in the after— noon versus in the morning. Neither the usage nor the value relationships were found to be significant, although going in the predicted direction (ru=.04, pu>.05, ru2=.002, rv=.05, pu>.05, ru2=.003) meaning that people who shopped in the afternoon more frequently than in the morning wished to use TPDS only a little more often and perceived TPDS as only slightly more valuable than morning shoppers. The fourth variable, shopping on weekends as opposed to weekdays, was also not significantly related to usage or value. Furthermore, the signs of these co— efficients portrayed relationships going in the opposite direction of that hypothesized (ru=-.O4, pu>.05, ru2= .002/rV=—.04, pv>.05, rV2=.002). These correlations indicated that people who prefer to shop on weekdays over weekends wished to use TPDS somewhat more frequently and perceived TPDS as being a bit more valuable. Summary of Test Results for Hypotheses 16-19’ Most of the relationships between current shopping or banking behavior and TPDS usage or value were found to be operating in the directions that had been expected meaning that people who tend to use more impersonal or less interpersonal communication oriented shopping or banking services tended to favor TPDS more (would use TPDS more often and are willing to pay more for TPDS) 146 than those using less impersonal or more interpersonal communication oriented services. One exception was the significant relationship found between shoppers who fre— quently drive to stores and TPDS usage. This relationship, although going in the opposite direction of that predicted, upheld the "if highly mobile, one wishes to travel less“ concept suggested by other hypotheses tests. It is interesting to note that people who reported more frequent use of impersonal shopping services were significantly more favorable towards the use of TPDS; while people having more frequent experience with imper— sonal banking services perceived TPDS as having a signifi- cantly higher value. People having frequent exposure to impersonal shopping and/or banking services could be expected to develop a certain expertise for handling such situations. Using a mechanical teller machine, in par— ticular, would acquaint people with the practice of making requests and directly interacting with a machine via pushbuttons. Reviewing Katzman's conclusion that the possession of abilities required to merely use a new communication technique increases the value of its output to the receiver; one could conclude that the impersonal banking behavior——TPDS value relationship supports Katz— man's findings, while the impersonal shopping behavior variable——TPDS usage relationship does not. The imper— sonal shopping behavior variable does, however, support 147 the more general concept that an innovation will be more readily adopted by those perceiving a lesser need for behavior change.5 The finding that frequent night-time shopping and TPDS usage was the only significant time of day or day of the week interpersonal shopping relationship is also important. Assuming most stores are open during the morning and afternoon of all weekdays and at least part of any weekend, then nighttime is when shopping behavior is most restricted. If frequent night shoppers have dif- ficulty shopping at other times, then they have the most to gain through access to TPDS. Yet, even these people were not willing to pay very much for TPDS. The fact that no significant relationships were found between any of the time of day, day of week variables and TPDS value may indicate that most people did not perceive access to TPDS as being very important to them, relative to their current shopping modes. That is, given a choice of driving to a store, or using TPDS, people did not perceive TPDS as being worth a very high monthly charge, if any charge at all. It is impossible to determine whether subjects actually con- sidered the cost of gas when answering the TPDS value question. It is quite likely that people, assuming they would use the more traditional shopping modes along with TPDS, viewed the monthly charge as an "extra" expense. 148 If so, the absence of any significant relationships between any of the time of day, day of week variables and TPDS value would be due to the fact that most people are not ready to substitute the use of TPDS for interpersonal oriented shopping modes. Multivariate Hypothesis Tests The following section presents the findings of the multivariate level hypothesis tests. As represented in Chapin‘s model, multiple regression was used to test clusters of independent variables against each of the dependent variable measures. In this situation, an acceptable research hypothesis was determined not only by significant relationships between the independent variables and both favorability measures (i.e., usage and value), but also required all of the independent variables to be significantly contributing to the regression equation in their predicted directions. Although some of the F-tests measuring the linear associations present in the regression equations proved to be significant (see Table 22), never were all of the independent variables contributing sig— nificantly in their predicted directions. Nor were the amounts of dependent variable variance explained by any of the regression equations appreciatively large. Thus, none of the multivariate level research hypotheses were supported. 149 TABLE 22 MULTIVARIATE HYPOTHESIS TESTS: SIGNIFICANCE OF REGRESSION EQUATIONS AND AMOUNT OF VARIANCE EXPLAINED Independent Variable TPDS Usage TPDS Value Class 2 2 F p r F p r Attitudes 1.85 .080 .072 .411 .895 .017 Demographics .586 .852 .029 3.47 .000 .149 Propensity to Engage 1.45 .104 .107 2.85 .000 .191 Opportunity to Engage 1.60 .140 .062 .637 .784 .026 Propensity and Oppor— tunity to Engage 1.70 L92 .165 2.37 .000 .217 Discussion of each hypothesis will therefore be directed towards examining trends in the amounts of variance explained and contributions of independent variables to the regression equations. As stated in Chapter III, Chapin's model suggests that there is inter— ‘ action occurring between the independent variables as one moves across the model from left to right. Given the use of a linear regression model, such interaction would be found wherever the amount of dependent variable variance explained by a total set of independent variables was greater than the sum of the variance explained by these variables when acting alone or in subsets. That is, the independent variables would not explain as much dependent variable variance when acting individually as they would when joined together in propensity or opportunity to 150 engage. Nor would propensity or opportunity to engage explain as much variation individually as they would when acting together. According to Sonquist's review of data analysis problems and additive versus interactive models in regard to multivariate model building, there are two major alternatives to the presence of interaction effects. If two or more independent variables had a fairly high level of intercorrelation, a certain amount of the variance explained by the best dependent variable predictor would be shared with the other variables. This sharing, some— times referred to as the substitution effect, would result in a decreased amount of variance explained as more and more intercorrelated independent variables were joined together in a linear regression. On the other hand, if no interaction or substitution was occurring, meaning that the independent variables were truly orthogonal, the sum of the amount of variance explained by any series of mutually exclusive independent variable groups would be exactly equal to the amount of variance explained when all of those variables were combined as a single group.6 If the variables used in this study are valid representations of the components of Chapin‘s model and if Chapin's model is a valid model of people's reactions to a new communication technology, such as TPDS, then the amount of variance explained will be expected to 151 increase as more independent variables are entered into the regression equation. The amounts of variance which were explained by the different groups of independent variables will be mentioned at each hypothesis test. The test findings will be followed by an overview of the entire model. The multiple regression statistics for each hypothesis test, including the beta weights and squared semipartial correlations of each independent variable, can be found in Appendix C. Hypothesis 20: People having a high propensity to engage will be more favorable towards TPDS than those having a low propenSity to engage. Propensity to engage, consisting of the combined forces of the attitudinal and demographic independent variables, explained 11 percent of the TPDS usage variable variance. Only two independent variables, attitudes towards technology (rsp2=.022) and interpersonal shopping orientation factor 5 (rsp2=.014), dealing with conver_ sations with salesclerks, were found to be significantly contributing to the equation in the predicted direction. The importance of attitudinal independent variables, in general, to determining TPDS usage levels was emphasized further when the total amount of TPDS usage variance explained by these variables was separated, as a group, L—l 152 from the demographic variables. The attitudinal variables, when regressed against TPDS usage (F=l.85, p=.08), explained 7 percent of the variance, while the demographic variables, as a group (F=.586, p=.852), accounted for only 3 percent. The need to evaluate demographic variables in addition to attitudinal variables was upheld by the propensity to engage-~TPDS value relationship (F=3.47, p=.000). Propensity to engage explained 19 percent of the variance in TPDS value. Here, seven independent variables, five of them demographic characteristics, were found to be significantly contributing to the regression equation. The amounts of variance explained, as expressed by the squared semipartial correlations, for those indepen— dent variables contributing in their predicted directions are shown in Table 23. Also shown in Table 23 are those contributing in the opposite direction of that predicted. ‘ Together these variables represented all but 1 percent of the variance explained by the total equation. Looking at the amounts of variance explained by the demographic variables (15 percent) as opposed to the attitudinal variables (2 percent) when regressed in separate groups against TPDS value; not only did the demographic variables explain considerably more TPDS value variance, their regression equation was actually significant (F=3.47, p=.000). While the attitudinal 153 variable regression equation was as bad as the demographic variable equation was in predicting TPDS usage (F=.4ll, p=.895). TABLE 23 VARIANCE EXPLAINED BY SIGNIFICANT INDEPENDENT VARIABLES Amount of Variable Variance Direction Explained Age 5.1% As predicted Lifecycle Status 3.9 As predicted Number of Cars in Household 3.9 Not as predicted Educational Status 2.2 As predicted Social Participation 1.8 Not as predicted Interpersonal Shopping Orientation, Factor 6 .7 As predicted Impersonal Shopping Orientation, Factor 3 .4 As predicted The two propensity to engage components thus have varying levels of prediction expertise, with attitudes related more strongly to TPDS usage and demographics with TPDS value. These trends contradict some of Katzman's iconclusions. Attitudes may reflect certain types or amounts of knowledge, thereby functioning as a component of informational status. If so, the relationship between attitudes and TPDS usage could follow Katzman's finding that higher informational status seems to be correlated with higher motivational levels to use new communication techniques. However, if one equates ability to first learn about an innovation (which Katzman attributes to 154 such things as education, mass media exposure, social participation, cosmopolitism, and interpersonal communi- cation skills) with informational status, then demographic characteristics would also be expected to be good pre- dictors of TPDS usage. Hypothesis 21: People having a high opportunity to engage will be more favorable towards TPDS than those having a low opportunity to engage. Opportunity to engage based upon interpersonal versus impersonal communication oriented banking and shopping behaviors, explained 6 percent of TPDS usage and only 3 percent of TPDS value. None of the independent variables contributed significantly to either regression equation. However, it is interesting to note which inde- pendent variables explained the most variance in each equation, given the bivariate level hypothesis test results associated with these variables. The largest amount of TPDS usage variance explained by a single independent variable was 2.3 percent, contributed by the interpersonal shopping variable concerned with travel to stores by car. As in the bivariate level test this variable was found to be inversely related with TPDS usage. The greatest contributor to TPDS value variance was the impersonal banking variable (1.1 percent). This variable was, as before, positively related to the 155 dependent variable. The shopping at night variable explained the second highest amount of variance for both TPDS usage (1.9 percent) and value (.5 percent). The impersonal shopping variable, as the remaining variable found to be significant at the bivariate level, explained the third highest amount of TPDS usage variance (1.3 per- cent). Thus, although no significant contributions to the multiple regression equations occurred, the variables found to be significant at the bivariate level were con— tributing the most to the multiple regression equation. Finally, considering the difference in the total amounts of variance explained by each regression equation, opportunity to engage seems to be a better predictor of TPDS usage than value. Again Katzman's conclusions are to some extent contradicted. Katzman‘s findings suggest that opportunity to engage, as a means of developing necessary skills, would be a better predictor of usage than value. Looking within the two regression equations, the same discrepancy between impersonal shopping and impersonal banking behavior that was discussed at the bivariate analysis level appears. That is, the impersonal banking variable conforms to Katzman's conclusions, while the impersonal shopping variable does not. l l . 156 Hypothesis 22: People having a high propensity and opportunity to engage will be more favorable towards TPDS than those having a low propensity and opportunity to engage. Significant relationships were found when all of the independent variables were regressed against TPDS usage and value. Yet, at this stage, where propensity and opportunity to engage were joined together in the pre— diction of TPDS favorability, the interaction of all inde— pendent variables did not contribute to any greater explanations of variance. The TPDS value equation explained 22 percent of the variance, thereby equalling the sum of propensity (19 percent) and opportunity (3 per— cent) to engage operating independently. The TPDS usage equation explained 17 percent, equalling the sum of pro- pensity (11 percent) and opportunity (6 percent) to engage when operating alone. These equalities represent the existence of orthogonal variable sets, and to that extent, support the use of an additive model in making predictions about TPDS favorability. An Overview of Chapin's Model The equalities found in the Hypothesis 22 test do not necessarily mean that no interaction or substitution effects have occurred anywhere within the model, but they do show that any such effects were cancelled out within the full regression equation. To look back through the 157 changes in variance explained at different stages in the model, Figures 3 and 4 were developed. Figure 3, working with just TPDS usage predictions, shows that the sum of the variance explained by the demo- graphic and attitudinal variables at the simple regression level (.09), explained 1 percent less variance than the sum of the variance explained by the attitudinal and demo- graphic variable clusters (.10), which in turn, explained 1 percent less variance than the single propensity to engage regression equation (.11). Examining just the amount of variance explained by the attitudinal (.07) and demographic (.03) clusters, as opposed to the sum l of the variance explained by either group at the bivariate level ( attitudes = .06, demographics = .03), interaction appears to have mainly occurred within the attitudinal variable set. However, there also appears to have been some interaction between the attitudinal and demographic variables as a set. 1 The sum of the TPDS usage variance explained by l l Ethe behavioral independent variables at the simple regres- ision level (.08) was decreased by 3 percent within the pportunity to engage regression equation. Therefore, a ubstitution effect occurred between some of the behavioral ariables when joined as a single predictor set. Figure 4, dealing with TPDS value as the dependent rariable, also indicates the presence of some interaction 158 .mCOHumsqm GOHmmmHmmH mHmHuHsE can mHMHHm>HQ an omens mama CH pwsHMmem mUGMHHm> .m .mHs mo. mmoH>HQzH mo 23m Noo. IIImAMUmez mumuw> mpcwxooz so monm Noo. IIImsHsHoz mumum> coocumumd CH mocm mmmmcm 09 mo. IIIIIIIIIIIIIIIIIIIIII uanz um monm muHssuHommo mo. llllllllllllllllllllllll use an d0hm maOHHUMmGMHB mcHxsmm HosowuomuousH oo. ,IlmCOHuoomcoHB OSchnm HmsomnomEH mo. mUHmmHQzH m0 29m Ho. UHonomdom CH mumu mo uwafisz OO. msnmum NHHHHnoz Haonsra NOO. ................... aonH>mHme anmo omen mom mo. vmo. mCOHHUMmGMHB msHmmonm HchmHomHmusH hIflmmMHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHW Ho. (JImCOHpommanB mCHmmocm HmGOmHoQEH voo. D mOHumHHmuomum Hoo. IIIIIIIIIIIIIIIIIIIIIIII MHcmz usHHm NH UHBQMHmo oo. IIIIIIIIIIIIIIIIIIII chmz ammopooum mo. .llvoo. . musmomxm MHpmz mmmz (llvoo. GOHummHOHuHmm HmHoom Ho. was woo. handpm mHomomMHH oo. msumum HMdOHuoosUm oo. smack mstHoz mmmmsm OB Hoo. mspmum UHEosoowOHoom Huang a co. xmm HH. mo. mflQDBHaed HdDQH>HQzH mo EDm OOO. uuuuuuuuuuuuuuuuuuuuuuuuuuu O uouoms Ho. IIIIIIIIIIIIIIIIIIIIIIIIIII m Houomm OHMmHMmuomumco No. lllllllllllllllllllllllllll H Houomm capsuHuud vmo. IlllllllldoHumucmHuo mCHmmocm HMGOmHmmHmusH . .||\I|..|-|I|I|I.III||.||\|.|‘|. NO. OO. uuuuuuuuuuuuuuuuuuuuuuuuuuu a Houomm oo. lllllllllllllllllllllllllll m uouomm Ho. IIIIIIIIIIIIIIIIIIIIIIIIIII N Houomm Ho. sOHumusmHHO msHmmonm HMGOmHomEH No. mmOHosnomB mounsou mmcsufluud .mCOHumswm COHmmmHmmH mHmHuHSE oCm mumHHm>Hn mo odHo> mode CH omCHMmem GOCMHHM> .a .OHC mo. mmOH>¢mmm HHQZH mo 23m Noo. Illmhmoxwwz mswuw> moCmMmmB Co monm mmcmCm OB moo. IIImCHCHoz mCmHo> COOCHmumm CH monm muHCsunommo Ho. IIIIIIIIIIIIIIIIIIIIII quHz um monm .mO. NOO. uuuuuuuuuuuuuuuuuuuuuuuu .80 F dorm NHo. mCOHuoomCmua mCHQQOCm HmCOmHmmHouCH Hoo. wCoHHUMmCmuB mCHmmOCm HwCOmHmQEH Noo. mCOHHOMmCMHB mCHxCom HMCOmHmmHopCH Ho. mCoHponanHB OCHmem HoComHmmEH mH. mUHmmémwozmQ H¢DOH>HDZH m0 25m No. oHosmmsom CH mumo mo Hmnfisz maHm> NOO. museum NnHHHnos HaOHmNra mo woo. nnnnnnnnnnnnnnnnnnn COHmH>mHoB lomo 03 NN. moHumHuwuomum oo. IIIIIIIIIIIIIIIIIIIIIIII mHowz uCHHm MN dHCmmnm Hoo. llllllllllllllllllll oHcoz unmoomoum mH.lllr moo. musmomxm MHUoE mmmz . Ho. fl, COHummHOHuHom HMHoom /fi©o mm¢ vo. nsumum mHo>omMHH mNo. mopoum HMCowumospm . CmEoz CHMHoz mmmmwmwe wm. msumum OHEOCoomOHMOm . oo. mm mH No. mMQDBHBB¢ HHDDH>HQZH mo EDm Ho. lllllllllllllllllllllllllll o uouomm NCO. lllllllllllllllllllllllllll m HOHUMW Hm . IIIIIIIIIIIIIIIIIIIIIIIIIII H Houowm HHMMMMWNMHCCU NHo. oo COHHMHCwHHO mCHmmonm HoCWmWMMMMWCH . .V\t\\\\\\\\\\\ . IIIIIIIIIIIIIIIIIIIIIIIIIII No Nmm. IIIIIIIIIIIIIIIIIIIIIIIIIII m Houomm oo. IIIIIIIIIIIIIIIIIIIIIIIIIII N Houomm NHo. COHumquHHO mCHmmonm HMCOmHomEH oo. hmOHOCaomB moHCBOH mwpsuHuus 160 and substitution. The sum of the variance explained by the attitudinal and demographic variables at the simple regression level (.20) was 3 percent greater than the sum of the variance explained by the attitudinal and demo- graphic variable sets (.17). The substitution appears to have occurred within the demographic variable set (.15) which explains 3 percent less than the simple regression sum (.18). An interaction effect (2 percent) was again found between the attitudinal and demographic variable clusters and propensity to engage. The behavioral independent variables, on the other hand, appear to be orthogonal. Summary of Findings The descriptive questions failed to reveal many major differences between how subjects used phone or mail order services. The only distinction found was concerned with where people learned about the products they ordered. Catalogs were the most frequently mentioned information source across both groups. However, people shOpping by mail appeared to rely upon a greater variety of infor- mation sources. This finding could be attributed to the existence of more mail order services than phone order services, and the variety of media chosen to promote such services. If so, it merely reflects the state of the art rather than any significant behavioral decisions. 161 Reactions to the concept of TPDS were fairly posi— tive in terms of service use, with 57 percent of the sample wishing to use TPDS often or very often. Per- ceptions of TPDS value were not as positive. Here, 25 percent of the sample was willing to pay over $10 per month for the service, but 28 percent were not willing to pay anything. When asked what type of product they would prefer to have access to via TPDS, groceries were the top choice of the sample. Reviewing all of the hypothesis tests, the inde- pendent variables values found to be significantly related to a desire to use TPDS more often were: 1. More positive attitudes towards technology 2. More positive attitudes towards interpersonal shOpping factor 1 (browsing, comparison shopping) 3. More negative attitudes towards interpersonal shopping factor 5 (talking with salesclerks) 4. Younger people 5. More cars in household 6. Higher frequency of shopping by car 7. Higher frequency of shopping at night 8. Conduct more impersonal shopping transactions While more positive perceptions of TPDS value were found to be related to: 162 More negative attitudes towards interpersonal shopping factor 6 (companionship) More positive attitudes towards impersonal shopping factor 3 (avoid salesclerks, already out) 3. Younger people 4. Earlier lifecycle status 5. Higher educational status 6. More cars in household 7. Lower social participation 8. Conduct more impersonal banking transactions Comparing these lists, the only independent variables found to be significantly contributing to the amount of variance explained for both dependent variable measures were age and number of cars in the household. Specifi— cally, the findings suggest that younger, more mobile people will be more favorable towards TPDS. In general, predictions of TPDS usage seem to be affected more by the attitudinal and behavioral independent variables than by demographic characteristics, whereas TPDS value predictions seem to be dominated more by demo— graphic independent variables. As suggested earlier, these findings do not completely support Katzman. If one may equate attitudes and behavior with informational status, then such variables would be expected to be 163 related to more frequent TPDS usage. However, Katzman's major contention is that certain demographic character— istics are related to being "first knowers." In this case, stronger relationships between demographic character- istics and TPDS usage should be found. The major variable Katzman relates to value per- ceptions is socioeconomic status. This variable did not prove to be significant. Katzman also relates "ability to use" with value perceptions. Experience with impersonal banking transaction could be enhancing one's ability to work with machines or impersonally handle financial matters, explaining the importance of that variable. The question is, where do the demographic characteristics fit into value perceptions? Looking at "ability to use" more generally as "if one needs something, they will make sure they are able to get it," it is possible that people who are: In an earlier lifecycle stage (lifecycle status, age) 2. Frequently mobile (cars in household) Generally busy (perhaps lower social participation) 4. Educated enough to understand how TPDS operates may perceive TPDS as a needed means of overcoming time and/or travel oriented restrictions. This perception coupled with impersonal communication oriented attitudinal 164 and behavioral preferences could reasonably result in a higher estimation of TPDS value. Indeed, the consideration of time or travel fac— tors appears to be an important issue in terms of general TPDS favorability. Trends within the data suggest that willingness to use TPDS more often and perceptions of TPDS as more valuable are related to: 1. Higher physical mobility status 2. More cars in household 3. More frequent travel to stores by car 4. More frequent night shopping 5. Less social participation 6. Less media use 7. Working women The first three of these independent variables indicate a lifestyle which may involve more travel. Shopping needs are responsible for at least some of this travel. The positive relationships between these travel character- istics and TPDS favorability suggest that people who travel a lot would like to travel less often. The need to shop at night, and lower levels of social participation and media use may be reflecting time restrictions. Unfor— tunately, it is impossible to determine from the data whether people reporting lower involvement levels are those who just do not desire, or get involved in, much social interaction or media use versus people having a 165 limited amount of time available. Nor is it possible to separate people who simply prefer to shop at night from those who have to. Stronger support for the time issue is, however, present in the fact that working women, who definitely experience time restrictions, did tend to favor TPDS. In terms of overall reactions to TPDS as an alter— native to travelling to a store to shop or buy, the con- flict recognized by Tauber is confirmed. That is, people having a more interpersonal communication orientation towards shopping were generally less favorable towards TPDS than those of a more impersonal orientation. The browsing, comparison shop factor, presented a most unsus— pected contradiction to this trend. Yet, this is not necessarily inconsistent with other findings of the study. If one regards browsing (particularly as it relates to a desire to compare product attributes or prices) as a high travel and time demand activity, then TPDS may be perceived as a means of fulfilling browsing needs which would definitely reduce travel and possibly save time. In this case, the browsing-TPDS usage relationship strengthens the position that travel and time factors are important to the TPDS situation. CHAPTER IV--NOTES lClyde L. Rich, "Is Random Digit Dialing Really Necessary?" Journal of Marketing Research 14 (August 1977): 305. 2Hubert M. Blalock, Jr., Social Statistics (New York: McGraw-Hill Book Company, 1972), pp. 296-97. 3Ibid., p. 285. 41bid., pp. 293, 297. 5Everett Rogers and F. Floyd Shoemaker, QQEEHBL cation of Innovations (New York: The Free Press, 1971), pp. 350—51. 6John A. Sonquist, Multivariate Model Building, Validation of a Search Strategy (Ann Arbor, Mich.: Institute for Social Research, University of Michigan, 1970); PP- 1-76. 166 CHAPTER V CONCLUSIONS AND RECOMMENDATIONS Background The purpose of this study was to explore how cer- tain human factors might affect the use of a single in- the—home telecommuting service, purchase—decisionmaking. Telecommuting technology allows physical travel, resulting in face-to—face interaction, to be replaced by technologi— cally mediated interaction via two—way communication sys— tems. A telecommuting purchase decisionmaking service (TPDS) was defined as a service by which manufacturers, wholesalers or retailers may provide information to their customers, allowing them to become aware of a product's existence on up to actually placing an order for a desired product or seeking decision reinforcement. In the past, much of the research conducted in regard to new communication technologies has been con- cerned with technological feasibility. Experience, how- ever, has shown that having a technology which is oper— ationally feasible is not enough to guarantee success in the marketplace. Indeed, Alex Reid, in a 1971 report to 167 168 the Sloan Commission, observed that telecommunication research should be shifted away from technological feasi- ibility towards the study of human aspects.l In a broad cross—cultural review of communication technology adoption, Katzman was able to identify higher and lower technological information status groups on the basis of certain demo- graphic characteristics. He concluded this work with the suggestion that the "equal access" approach to reducing this gap has failed because motivational factors as well as other aspects of the psychological environment have not been considered.2 The development of telecommuting technology has been accompanied by some empirical research dealing with teleconferencing or computer conferencing, as they relate to business operations and working conditions.3 The psy- chological or behavioral reactions of individual consumers to nonoccupational, in—the—home telecommuting services have not received much attention. A purchase decision— making service was singled out for study because retailing and banking industry trends seem to support more impersonal communication oriented transaction modes; and if adopted, TPDS could supply the high volume of use required for interactive cable television system growth. The primary issue raised by the prospect of using TPDS was believed to be whether people would be willing to use an on—demand impersonal communication oriented 169 service, as opposed to physically travelling to a store. This issue was attributed to the idea, presented by Tauber, that people might not go to stores just to buy products. That is, certain social motivations may be satisfied by physical travel and contact with in—store stimuli.4 If so, the use of TPDS would prevent people from achieving such satisfactions. A potential solution put forth was that people who lost, or were unable to use, this source of social motivation satisfaction would become more involved with other social activities. It was assumed that the extent of this issue would vary across different types of people. A general systems model developed by Chapin, in his work concerning the behaviors of population aggre- gates in relation to urban planning, was used in this study to explore how different types of people might react towards TPDS. Chapin attributes any decision regarding activity participation to the joint action of two factors-—propensity to engage and opportunity to engage. In accordance with Chapin, propensity to engage was viewed as consisting of the interaction between atti- tudinal variables (which predispose individuals to action) and demographic variables (which precondition actions). Opportunity to engage is defined by Chapin as the inter— action between individuals' perceptions of their ability to have access to necessary facilities or services and 170 the quality of those facilities or services.5 Given that TPDS was not available to consumers in the sample, the opportunity to engage measure was represented in this study by use of currently available impersonal communi- cation oriented shopping and banking transactions. This substitution was made on the assumption that peoples' perception of TPDS, in terms of the effort involved to gain access (availability) and performance standards (quality), might be affected by their past levels of participation in activities having certain characteristics in common with TPDS. The independent variables in this study consisted of attitudinal, demographic, and behavioral character— istics. Due to the exploratory nature of the study, the specific characteristics measured were drawn from pre- vious research in the fields of communications, transpor— tation, sociology and marketing and the retailing and banking industries. Following the findings of Nilles et al., concerning attitudes towards a number of telecommuting services,6 the dependent variable, favorability towards TPDS, was split into two projective measures, potential usage (how often one would use TPDS if it was available for free) and perceived value (how much one would be willing to pay for access to TPDS on a monthly basis). These questions were asked following the presentation of a description of how TPDS would operate. 171 Information regarding how people currently use phone and mail shopping services was also collected. It was believed that such information might provide further insight into how people might be expected to use TPDS. The data were collected from 261 adults in the East Lansing, Michigan area via a telephone survey. This sample was found to be slightly biased towards younger, more highly educated people of a higher socioeconomic status. The sample also contained more females than males. Given that many of the phone calls were made in the daytime, this sample is not necessarily unrepresenta— tive of the East Lansing area population. The sample is not, however, adequate for making nationwide generaliza- tions. A more serious problem posed by the sample biases is their effect on the hypothesis tests. The hypotheses dealing with demographic independent variables clearly suggest differences between higher and lower variable values. If a demographic independent variable has a skewed distribution, the range of variable values, and variance within that range, may be decreased. This narrowed variable may not be adequate to properly test the related hypotheses. That is, the chance of finding significant differences could be greatly reduced by work— ing with a lower level of variable value variation than that specified in the hypotheses. Specific demographic 172 quotas or a stratified random sample could be used to eliminate this problem in future work. Conclusions People were fairly positive towards the idea of using TPDS if the service was available to them for free. Yet, people did not seem to consider TPDS to be particu- larly valuable to them. Groceries were cited more often than any other type of product as what people would like to have available to them via TPDS. Thus it appears as though a service offering groceries on a low cost basis would be well received. Looking at how people currently use impersonal shopping services, about two out of every five people were found to have shopped by phone or mail in the three months preceding the interviews. Statistical analysis failed to reveal any outstanding differences between heavy and light users or the use of phone versus mail order services. While these people did not necessarily depend heavily upon mail or phone shopping services, they did appear to be habitual users. The significantly posi- tive relationship found between impersonal shopping ser- vice use and potential TPDS use, suggests that these people could be expected to be fairly receptive towards TPDS. Frequent users of impersonal banking services were found to perceive TPDS as significantly more valuable 173 than less frequent users, rather than favoring TPDS usage. It is difficult to access why this distinction occurred. It may be that people are more accustomed to paying for banking services than shopping services. Thus, people who currently use impersonal shopping services were pleased with the prospect of more shopping opportunities, but did not necessarily want to pay for it. On the other hand, people who currently use impersonal banking ser— vices may not be ready to use TPDS as frequently, but were able to perceive it as worth paying for. Perceptual differences such as these may help to explain why Nilles et al. found discrepancies when they asked people to rate a group of potential TPDS services in terms of value and desirability. Reviewing the entire range of hypothesis tests, there are several general conclusions which may be drawn. 1. The communication—transportation trade-off potential of TPDS is a major factor in determining favora- bility towards TPDS. More mobile people may be expected to be more favorable towards TPDS, and perhaps any tele- commuting service, due to its capacity for travel reduction. While, with the possible exception of people whose travel is restricted by the presence of children, people who are less mobile may be expected to be less favorable towards TPDS. It may be that such negative reactions stem from the need for social interaction, i.e., social motivation 174 satisfaction. This would confirm the existence of communi— cation oriented obstacles to telecommuting acceptance. 2. People who have more negative attitudes towards in-store interpersonal conversations may be more favorable towards TPDS. People do associate varying levels of inter— personal and impersonal communications with store related activities. In the words of Tauber, there is a distinction between shopping and buying. Shopping, in the form of browsing (where peOple are primarily concerned with com— paring product attributes and prices), may be easily con— ducted via TPDS. However, people who value discussions with salesclerks or others while in a store, may be expected to be more negative towards TPDS use. While, people who generally seek to avoid salesclerks or the company of others while shopping may be expected to place a higher value upon TPDS. 3. People who are less active socially place a higher value upon TPDS. The degree to which people inter— act with others varies greatly. Some people, experiencing time constraints, may be responding positively due to the greater flexibility TPDS offers. The more impersonal communication orientation may be appealing to those who Simply prefer to avoid social interaction. As before, this situation may be generalized to any telecommuting services which entail similar losses of social interaction. 175 4. Age may be the best demographic predictor of favorability towards TPDS. Younger people may be expected to use TPDS more often and to pay more for TPDS than older people. It is difficult to understand why age was found to explain more variance than any other demographic variable. One possible reason is that the college student members of the sample, who may experience certain travel or time limitations, biased the age variable. Lacking previous research concerning TPDS, the reliability of these conclusions is questionable at this time. The first conclusion displays a concern over travel and time activity factors. Though society has always tended to strive towards improving travel conditions and decreasing the amount of time needed to conduct activities, the average consumer of today may be more aware of these needs, particularly in regard to the energy crisis. The second conclusion confirms the idea that people experience more than just a "product purchase" motivation for going to a store. When the second and third conclusions are viewed together, the possibility that trade-offs between more and less socially interactive activities could com— pensate for the impersonality of TPDS appears dim. The fourth conclusion is in contradiction with the work done by Katzman. While age might be considered a variable contributing to higher informational status and therefore 176 correct in predicting TPDS usage, Katzman clearly con— cluded that economic status was the major contributor to value perceptions. Limitations and Contributions There are several limitations to this study. The number of independent variables (26) used in this study is rather large relative to the sample size (261). Under such conditions, multiple regression analysis will tend to produce an inflated multiple correlation. This means that the total amount of variance explained by a new set of data, if the regression equations derived in this study were applied, would be lower. These regression equations, themselves, may be unreliable because the beta weights are also affected by the number of variables to sample size ratio. The reliability of the beta weights is further reduced when any of the independent variables are interrelated. While, in this study, the independent variables were not found to be highly correlated among themselves, there is reason to believe that some substi— tution and interaction effects did occur at certain levels of analysis. Such effects can be expected to have affected both the beta weights and the multiple cor— relations of the regression equations:7 To obtain more reliable findings, a larger sample and orthogonal independent variables would be needed. It may, however, be impossible to achieve complete 177 orthogonality among the types of variables used in this study. Reliability might then be increased by reducing the number of independent variables used, with this group of variables being as orthogonal as possible. Given Chapin's model and the literature review from which the independent variables of this study were drawn, all twenty—six variables were deemed important. Thus, none of them were dropped while testing Chapin‘s model. How— ever, if the goal of identifying a minimum number of inde- pendent variables which will account for a maximum amount of dependent variable variance were to be adopted, then some of the independent variables should be deleted. Kerlinger and Pedhauzer suggest two criteria upon which independent variable deletion may be based: significant contributions to the total amount of variance explained and meaningfulness. This second condition being defined as whether the increment added to the squared multiple correlation is substantial relative to the research situation, such as the effort involved in obtaining the additional data.8 This study utilized standard or forward solution regression analysis without a specified order of entry. No particular order was designated because neither Chapin's (model, nor the literature reviewed, suggested any basis for making such a decision. This technique enters inde- pendent variables into the regression equation in rank 178 order of their ability to explain variance in the depen— dent variable, after partialling out all variables already in the equation. Application of the variable deletion criterion of significance would involve examination of the F tests calculated on each independent variable as they were entered into the regression equation. All significant variables were listed in Chapter IV and were noted by an "*" in Appendix C. Given the lack of complete orthogon- ality between independent variables and the sampling problems discussed earlier, it is difficult to determine if these significant variables are truly the best pre- dictors. Even if the sample had been large and represen- tative enough to be considered fairly reliable, confusion would still arise in regard to potential changes in the amount of variance explained by a variable depending upon whether it was added to the regression equation before or after other variables with which it is correlated. To avoid this problem of order effects, regression analy— sis would be needed on every possible combination of the independent variables. Comparisons made across the con— tributions of each variable in each situation would then provide a clearer picture of the relative values of variables. Aside from analysis issues, the study results may have also been hindered by measurement validity problems. Although an attempt was made to develop an easily 179 understood description of TPDS, it is impossible to deter— mine whether the concept of TPDS was accurately perceived by subjects. Nor is it possible to determine how inac- curate perceptions may have affected responses to the dependent variable questions. This sort of controversy is an inherent part of any concept test or forecasting technique. The only way one can improve the validity of such measures is to use subjects who are more likely to understand the concept. The fact that socioeconomic status failed to con- tribute significantly to the regression equation, particu— larly in regard to TPDS value, may also stem from a validity problem. The socioeconomic status measure may have been distorted when the status of students was rated according to the highest wage earner in their family, while nonstudents were rated according to the highest wage earner of their household. In terms of the use of Chapin's model, the vari— ables constructed as measures of opportunity to engage may have been improper representations of that model com— ponent. A situation where people actually had access to TPDS, or any other telecommuting service, would be required to measure opportunity to engage in that service as defined by Chapin. Some of the other variables such as physical mobility and social participation may have also been f“ my *— a” 180 improperly measured. Here, discrepancies could exist between the frequencies reported by subjects and actual behavior. The best, but next to impossible, means of measur— ing such behavior would have been direct observation such as that used in anthropological studies. Behavior diaries, which are somewhat easier to administer, may have produced more accurate information in regard to these variables. Lastly, the length of the questionnaire may have disturbed some subjects. However, outside of the rejection of several incomplete surveys, it is difficult to determine the extent to which this may have affected the findings. Suggestions for Future Research Given the problem in measuring the opportunity to engage component of Chapin's model, it may be more appro- priate to either use a different model or drop the oppor— tunity component of the model in future research. If the opportunity component were to be retained, a more repre- sentative measure may be found as two-way cable systems begin to offer interactive programming meaning that while opportunity to engage in TPDS use might not be measured, opportunity to engage in some form of telecommuting could be. The TPDS use--attitudinal independent variable, TPDS value-—demographic independent variable relationships found in this study, may allow future research to be more 181 specialized. If research were to be directed at usage or value issues on an individual basis, the use of attitudinal or demographic independent variables might be expanded or reduced accordingly. The major methodological limitations of this study could be eliminated through the use of behavior diaries in combination with telephone or personal interviews with a larger, more representative sample of people, who know more about interactive cable television. For example, a sample drawn in Columbus, Ohio where a two-way cable television system recently started programming which per— mits viewer—system interaction could lessen the problem of dependent variable validity. Future research could be strengthened further if the sample was proportionately stratified on the basis of some of the independent variable measures. Particularly important would be some of those which were found to be operating in the opposite direction of that suggested by previous research. This would apply to such variables as physical mobility, social participation, some forms of media use and the browsing shopping orientation. If a larger more representative sample was obtained, a more reliable analysis of the relative values of the independent variables could be achieved. Once the list of independent variables was reduced, further repli- cation or at least cross—validation would be desirable. 182 Having strengthened the overall measurement pro— cess, research should be expanded to cover alternative telecommuting services. Study of the human factors impor— tant to a multitude of service areas might uncover a set of variables which could be used to assess the impact of telecommuting in general. Such variables would be impor- tant not only to the cable television industry but also to such fields as transportation, marketing, advertising, sociology, and psychology in adapting to fit the future needs of society. CHAPTER V--NOTES lAlex Reid, New Directions in Telecommunications Research (a report prepared for the Sloan Commission on Cable Communications, 1971), P. 3. 2Natan Katzman, "The Impact of Communications Technology: Promises and Prospects," Journal of Communi- cation 24 (Autumn 1974): 47—58. 3Lesley A. Albertson, "Telecommunications as a Travel Substitute: Some Psychological, Organizational and Social Aspects," Journal of Communication (Spring 1977): 33. 4Edward M. Tauber, "Why Do People Shop?" Journal of Marketing 36 (October 1972): 48. 5F. Stuart Chapin, Jr., Human Activity Patterns in the City, Things People Do in Time and Space (New York: John Wiley & Sons, 1974), pp. 32-35. 6J. M. Nilles, F. R. Carlson, Jr., P. Gray, and G. J. Hanneman, The Telecommunications-Transportation Tradeoff: Options for Tomorrow (New York: John Wiley & Sons, 1976), pp. 124-25. 7Fred N. Kerlinger and Elazar J. Pedhazur, Multiple Regression in Behavioral Research (New York: Holt, Rine- hart & Winston, Inc., 1973), p. 282. 8Ibid., pp. 70-75. 183 APPENDICES APPENDIX A SURVEY QUESTIONNAIRE APPENDIX A SURVEY QUESTIONNAIRE Hello. Ity name is . One of my classes at / Michigan State University is conducting a research study concerning h-V shopping behavior. Your household was selected at random to parti- cipate in this survey. May I ask you some questions? I. Do you ever shop for clothing or nongrocery household items for yourself or your household? YES SOMETIMES NO IF NO: I'm sorry, I need to talk to people who do shop for clothing and nongrocery household items, thank you for taking the time to talk to me. Good-bye. People shop for clothing and nongrocery household items in alot of different ways when they go to the store. I'm going to read to you a list of some of these ways and I'd like you to tell me how often you shop in that fashion. Please tell me how often in terms of very often, often, not often, not very often, or never. 2. How often do you go shopping with others? V0 0 NO NVO N (85 3. How often do you shop around at different stores to compare merchandise and prices before buying things? V0 0 NO NVO N (9) 4. How often do you go shopping at night? VO 0 NO NVO N 10 5. How often do you spend time just browsing . through stores without buying anything in particular? V0 0 NO NVO N (115 6. How often do you go shopping on weekends as V0 0 NO NVO N opposed to weekdays? (:57 7. How often do you spend time looking for bargains V0 0 NO NVO N or SPeCial sales when you are shopping? 'fijj 8. How often do you go to the store before or after V0 0 NO NVO N being somewhere other than in your home? . (1:7 184 185 9. How often do you buy things you didn't intend to buy when you started out on your shopping trip? V0 0 N0 NVO N (155 10. How often do you talk to sales clerks about things other than the merchandise you're looking at? V0 0 N0 NVO N (155 11. How often do you go shopping in the morning as opposed to the afternoon? V0 0 N0 NVO N (175 12. How often do you have difficulty deciding what to buy while in the store? V0 0 NO NVO N - (185 13. How often do you use a car in order to go shopping? V0 0 NO NVO N (195 14. How often do you talk about your shopping activities with friends or relatives? . VO 0 NO NVO N (205 15. How often do you go shopping for just one or two items rather than waiting until you need several things? V0 0 NO NVO N (215 There are some other aspects of shopping which are particularly important to some people but not of much importance to others. I'd like to know how important these things are to you in terms of always being important, sometimes important-depending on how much time you have or never important. 16. How in ortant is it to you to shop quickly? AI SI NI P (22) 17. How important is it to you that you get everything AI 51 N1 you need to bu in one store? y (235 18. How important is it to you when choosing which store to go to that it be the closest one which carries what AI SI NI 9 you need to buy. 24 19. How important is it to you that sales clerks I approach you and offer assistance? AI SI N 25 ’ d any 20. Have you ever shopped by telephone, that 1s ordere nonfood or drug items over the phone within the last 3 months? YES NO IF NO: GO TO 026 IF YES: How often would you say you shop by phone? V0 0 NO NVO N 2 186 21. Where do you usually obtain information about the items you purchase by phone? (MARK FIRST MENTION=1, OTHERS=2) NEWSPAPERS__ MAGAZINES__' CATALOGS__ MAIL CIRCULARS__ TELEVISION__ RADIO__ PEOPLE__ OTHER__ 22. Do you usually use delivery services if they are available? YES NO 23. Considering all of the items you've purchased in the last three months, where would you say you shop most often? IF THIS IS NOT A STORE (ASK IF YOU ARE NOT SURE): GO TO Q24 IF THIS IS A STORE: How long would it take you to travel from your home to this store? LESS THAN 15 MIN__ 15-30 MIN__ 31-45 MIN_ 46-60 MIN___ MORE THAN 60 MIN__ 24. What was the last thing you purchased by phone? (IF MORE THAN ONE TAKE FIRST MENTIONED) 25. How often do you plan to shop by phone in the future? V0 0 N0 NVO N 26. Have you ever shopped by mail, that is ordered any nonfood or drug items through the mail within the last 3 months? YES NO IF NO: GO TO Q31 hop by nail? ' : ould you say you s_ . . IF YES How often w V0 0 NO NVO N 27. Where do you usually obtain information about the items you Purchase by mail (FIRST MENTION=I, OTHERS=2) NEWSPAPERS MAGAZINES___CATALOGS___MAIL CIRCULARS__> TELEVISION___RADIO__.PEOPLE___OTHEE__ 28. Considering all of the items you've purchased in?the last three months, where would you say you shop most often. 3% 4 J 35 (337 (777 _ / (573797 (7707 187 IF THIS IS NOT A STORE(ASK IF YOU'RE NOT SURE): GO TO Q29 IF THIS IS A STORE: How long would it take you to travel from from your home to this store? (415 LESS THAN 15 MIN__ 15-30 MIN__ 31—4iMIN__ 46-60 MIN__ MORE THAN 60 MIN__ 29. What was the last thing you purchased by mail? (IF MENTION MORE THAN ONE ITEM TAKE FIRST MENTIONED) (425 30. How often do you plan to shop by mail in the future? VO 0 NO NVO N (43} 31. It has been suggested that how people shop is related to other activities in their daily lives, so now I'd like to ask you about some Of your daily activities. Could you tell me how often you do the following things? Again please respond in terms of very often, often, not often, not very often, or never. 32. Attend parties or other organized social functions? VO 0 NO NVO N 33. Attend church or participate in church related activities? V0 0 N0 NVO N 34. Attend cultural activities such as movies, theater, visits to museums, etc.? V0 0 NO NVO N 35. Attend sports activities like watching a game or tournement? ' V0 0 N0 NVO N 47 6. Pa t' i t in club or or.anization activities? 3 r 1c P3 e 8 VO 0 NO NVO N 37. Participate in cultural actiVities, such as performing in concert or giving a lecture? V0 0 NO NVO N JJJflfiéllfiilfi 38. Participate in individual sports such as golf, tennis, bike riding? VO 0 NO NVO N O 39. Participate in group sports like football, baketball or baseball? V0 0 N0 NVO N VO 0 NO NVO N \n F“ 40. Participate in volunteer work? VI N 41. Talk with members of your household? VO 0 NO NVO N \II \A) 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. home? 188 Write letters to friends or relatives? Visit with friends or relatives over the phone? V0 0 NO NVO N V0 0 NO NVO N Visit informally with friends or relatives in person? Read a magazine or Travel or tour? book for pleasure? Drive or ride around strictly for pleasure? Just sit and think? Work alone on hobby or craft activities? Work with others on hobby or craft activities? Talk frequently with others while involved in any of the group oriented activities previously mentioned? V0 0 N0 NVO N V0 0 NO NVO N V0 0 NO NVO N VO' 0 NO NVO N V0 0 NO NVO N VO 0 NO NVO N V0 0 N0 NVO N VOONONVON Thinking about all of the trips you have made anywhere during the last week, how often did you travel a distance within 5 miles of your What about further IF NEVERzGO TO Q56. IF NOT NEVER: How 10 miles from your IF NOT NEVER: How 30 miles from your IF NOT NEVER: How 50 miles from your IF NOT NEVER: How than 5 miles from your home? often did you travel further than home? often did you travel further than home? often did you travel further than home? often did you travel further than 100 miles from your home? VO 0 NO NVO N VO 0 N0 NVO N V0 0 NO NVO N VO 0 NO NVO N V0 0 NO NVO N V0 0 NO NVO N IF NOT NEVER: Did you ever travel further than 400 miles from your home during the last week? YES NO _._—-__— \n QAAAWJJWAQ Vi 001”] ififififiifi a 189 IF NEVER: GO TO 853 53. Imagine being able to do all your shopping without ever leaving your home. Through the use of a push button console attached to your television set, you could "call up" any store and view products and product information right on the tv screen. You could browse through a vast assortment of products, or even place an order and arrange for delivery of the products right to your home. This service would be available 24 hours a day, every day of the week. 54. If this service was going to be offerred only for certain kinds of products, what type of product would you most want to have made available in this fashion? (715 55. If such a service was available to you for free, how often do you think you would use it? VO 0 NO NVO N I7§5 56. If you had to pay a monthly fee for access to such a service what's the maximum amount you would be willing to pay within a / 0 to $50 range? $ 73-4 There is also some reason to believe that there are similiarities between how people shop and their banking behavior. So now I'd like to / / / / ask you some questions about how you use bank facilities. (1 - 5) 57. Do you remeber seeing or hearing any advertising for a bank recently? YES NO IF NO : GO TO Q58 IF YES: Could you tell me what bank it was for and/or where you saw or heard it? BANK RAD TV NP MAG OUTDOOR D.MAIL / / —- — — —- -—- — (‘74)) / / _- _. _... __ —— — (lo-12) 58. Thinking about where you do most of your banking, are there any sPecial services available, such as mechanical tellers, mail or phone transactions? ($37 MECHANICAL TELLER PHONE MAIL___ IF MENTION MECHANICAL TELLER: Do you have a mechanical teller card? YES NO (14) __——- ——— 190 IF DON’T MENTION MECHANICAL TELLER: Do you know if any banks in this area have mechanical tellers? YES____ NO —. iF NO: GO TO Q59 IF YES: Can you tell me which banks there are and/or what they call their mechanical tellers? ELSB—TELLER 24_ ELSE—READY TELLER ABM-READY TELLER_ AB&T-TFLLER 24_ OTHER 59. How often do you use the mechanical teller, phone banking or banking by mail services available to you? / / {17-195 MECHANICAL TELLER VO 0 NO NVO N PHONE V0 0 NO NVO N MAIL VO 0 NO NVO N IF DO NOT REPORT USE OF MECHANICAL TELLER: GO TO S63. IF D0 REPORT USE OF MECHANICAL TELLER: CONTINUE WITH NEXT QUESTION: 60. Thinking about the various transactions you conduct via the mechanical teller, what kind of transaction do you perform most often? Are there any available to you that you would never want to use? (RECORD AS FOLLOWS: l=MOST OFTEN, 2=NEVER, DO NOT AID THEM BY READING THE CATAGORIES) . WITHDRAW CASH FROM CHECKING ACCOUNT TRANSFER $,C TO S WITHDRAW CASH FROM SAVINGS ACCOUNT TRANSFER $,S TO C / / WITHDRAW CASH FROM MASTER CHARGE TRANSFER $,MC TO C DEPOSIT TO CHECKING ACCOUNT PAYMENTS OF ANY KIND ____PEPOSIT TO SAVINGS ACCOUNT OTHER ‘ 191 61. Which of the following locations best describes the mechanical teller machine that you use most often? NEAREST YOUR HOME NEAREST WHERE YOU WORK NEAREST WHERE YOU SHOP BETWEEN WORK AND HOME___ BETWEEN SHOPPING AND HOME 62. If you were to move to another city, would you seek a bank offering a mechanical teller service? YES___ NO___ 63. Due to the fact that many banks and other businesses have been mechanizing their services, such as providing mechanical tellers for customer use, many opinions have been expressed by people concerning the impact of technological developments upon their lives in general.-I'd like to know if you strongly agree, agree, feel neutral, disagree, or strongly disagree with some of these statements. a. People have become too dependent on machines. SA A N D SD b. People shouldn't worry about harmful effects of technology, because new inventions will always come along to solve problems. SA A N D SD 0. It would be nice if we would stop building so many . machines and went back to nature. SA A N D SD d. Technology has made life too complicated. SA A N D SD 64. Getting back to your banking activities, does your bank have a drive—in window with human tellers working at it? YES NO IF YES: How often do you use the drive—in window: V0 0 NO NVO N 65. How often do you go in the bank to carry out your banking transactions? V0 0 N0 NVO N 66~ When you actually travel to the bank, how often do you combine going there with other activities? V0 0 NO NVO N 67. How often does anyone go to the bank with you? V0 0 NO NVO N IF VERY OFTEN OR OFTEN: Who usually accompanies you? CHILDREN OTHER RELATIVES__ FRIENDS“ 33 fl 68. 70. 71. 72. 192 Now I'd like to ask you some questions about the last time you actually went to the bank. When was the last time you went to the bank? THIS WEEK: M T W TH F S SU LAST WEEK: M T W TH F 5 SU ON LAST PAYDAY M T ’w TH F S SU WHICH WAS OVER 2 WEEKS ACO I'cEICH. WAS OVER 3 WEEKS ACO : ANY OTHER OVER 2 WEEKS AGO __ Did you go inside, use a drive-in window, or a mechanical teller? INSIDE DRIVE—IN___ MECHANICAL TELLER__ (IF THEY USE)" A MECHANICAL TELLER AT A DRIVE—IN CHECK BOTH) IF INSIDE OR DRIVE-IN: Did you make a transaction which cchld only be conducted at the drive-in window or inside? YES NO Were you satisfied with the service you received? YES NO 'Do you remember what time of day it was? EARLY MORNINC__ MORNING___ AFTERNOON LATE AFTERNOON (12:01-8am) (8:01—noon) (12:01-4pE7— (4:01-6pm) " EARLY EVENING EVENING LATE EVENING CAN'T REMEMBER (6:01—8pm) ( 8:01-1fi) (lO:Ol—MidnigH€) Which of the following banking facilities were available to you at this time? DRIVE-IN INSIDE MECHANICAL TELLER___ MT AT DI IF MORE THAN ONE FACILITY WAS AVAILABLE: Thinking about when you arrived at the bank, which one of the follow1ng bank facilities had less people waiting in line? (USE ONLY THOSE PAIRS REFERRING TO WHAT WAS AVAILABLE TO THIS PARTICULAR RESPONDENT) AT THE DRIVE-IN WINDOW OR INSIDE THE BANK AT THE DRIVE-IN WINDOW OR AT THE MECHANICAL TELLER AT THE MECHANICAL TELLER OR INSIDE THE BANK (397 (757 (47.7 T257 W T477 193 Lastly I'd like to ask you a few questions about yourself and your household. This information will be used just to describe as a group / / / / the kind of people who have answered these questions. I‘i:3?'7““ 73. During an average day, how many hours do you personally spend listening to the radio: IN THE MORNING ,IN THE AFTERNOON.___JAND IN THE EVENING . / / / / / 74. During an average day, how many hours do you personally spend watching television: IN THE MORNING ,IN THE AFTERNOON___3AND IN THE EVENING . / / / / / (l2—l7) 75. Is cable tv available in your neighborhood? YES . NO (135 IF NO: GO TO IF NO BELOW IF YES: Do you subscribe to cable tv? YES NO ( 19 3 IF NO: Have you ever subscribed to cable tv? (205 YES NO 76. How much time do you personally Spend reading newspapers during an / average day? MINUTES (21-22) 77. How many magazines do you read on a regular monthly or weekly basis? . / 23-24) 78. What's your marital status? SINGLE MARRIED DIVORCED/WIDOWED/SEPARATED____ "”‘ ““” (255 79. How many people live in your household? l___ 2-3 4-5 6—7 over 7 (557 IF MORE THAN 1: Are any of them children under 13 years of age? YES NO (275 _-—-— Are any of them children between 13 and 18 years old? YES NO ___—.— Ea 194 IF 1 OR MORE AND SINGLE/DIVORCED/WIDOWED/SEPARATED: What is your occupation? IF STUDENT: Thinking about your family what's the occupation of the highest wage earner? / / (29-315 IF MORE THAN 1 AND MARRIED: What is the occupation of the highest wage earner in your household? IF SPEAKING TO A FEMALE: How many hours per week do you spend working outside your home? / (32-337 80. What was the last year of school you attended? COLLEGE GRAD SOME COLLEGE HIGH SCHOOL GRAD SOME HS '_— (7.7 81. How old are you? UNDER 18 18-24 25-30___ 31—34___ 35-45~__ 46-55___ 55—65 OVER 65 (355 82. How many cars are owned by members of your household? 1 2____3____4___ 5___ OVER 5 (365 33. Are all or some of the banking accounts you have with a bank, 7 credit union, or savings and loan company in the LanSIng-East Lansing area. YES_____ NO___ C377 IF NO: GO TO THANK YOU ..... IF YES: Could you please tell me what kinds of accounts you have and where they are located? BANK, CREDIT UNION 0R SAVINGS AND LOAN ACCOUNT TYPE LESS THAN 1? / (BB-415 / / / 42-45 / / / (43-49) / / / —5 84. Have you had any of these accounts for less than one year? (”IDENTIFY WITH x) 85. How many years have you live in the Lansing—East Lansing area? S THAN 1 YEAR 1 2 3 4 §____6____7____OVER 7___ LES '— "'" ‘— — _' ("527 THANK YOU VERY MUCH FOR TAKING THE TIME TO ANSWER THESE QUESTIONS FOR ME. GOOD-BYE. T: MALE FEMALE RECORD SEX OR RESPONDEN ___ ___ (337 APPENDIX B INTERITEM AND INTERVARIABLE CORRELATION MATRICES mm.- Mama muflquunmm BOUQHB cfllm>HMQIIHm Nm Hm EmuH ma. HHh NH. mo. no. wH.I mmHHE oov Allvm mm. mm. mm. mm. mo. mmHHE OOH Allmw mm. MN. mm. No.1 meflE om Alnmm ow. mm. Ho. meflE om Allaw mw. mo. meHE OH Allow ma. mmHHE m Allmm mmHHE m CH\3IImm mm .Mm mm mm mm mm mm amuH 6 9 wamom hwflaflnoz Hmoflmwsm 1 mm. ma. mw. QMHH Uwumoflamfioollmm ma. mv. wusumc ou cusuwmllmm mm. augeg Ufisonmnunm pcmwswmmw OOBIImm mm Mm Mm mm EmuH wamom hmoaoczome mUHmBOB wwwsuflupm mmMDmdmz HHQZH mo quom mqm wm QMmHZdmeIImEZMHUHmmmOU ZOHBflqmmmoo EMBHmMBZH m NHQZWAmd ma. aflma an mogmunom mcoam an monmllom om mm EmuH mamom maflmmozm HMGOmHmmEH ma. NH. ma.l mwmwxmwB Hw>o mwcmxowz Go mogmllm mm. HH. mGHCHOE Hw>o coocuwwmm ma monmllaa HH. #gmflc “m moamnuw 197 Hmo mg mafimmocmllma 9 mm v mm sme meSmmmz mcflmmonm chomMmmHmucH mw. mo. HHmzllmn oo.! wcosmllmn Hmaamu Havacmaomzllhn NM mm MM EmuH mamom mcfixcwm HMGOmHmmEH 198 NN. hN. av. vm. HH.! om. mH. NH. ON. ON. 0H. MN. Ho. 5N. mo. mN. meow GOHummHoHuumm HMHOOm O N 0 mo. 50. mH.I NH. oo. wo.l $0.! we. mo.l mo.| No.1 l on m - vm. NN.I No. no.l oo.! NH.I mo. mo. mH.I mo.l ho.l hm. HH.I mo.l wo.| MH.I v0. mo. mo.l VN.I mH.I wo.l oo. mo.| hH.I Ho.l OH.| No.l mo. wN.I mo.l no.l oH. mo. mo. no. 0H. mo. ON. OH. m2. FH. mo. 0H. mo. mN. NN. HN. Nm. VH. NN. Ho.| mo. mH. HH. HH. mo. mo. vo. mo. NH. mH. mH. mH. HN. no.l wH. HH. vH. ON. mo.l mm. mm. vN. MN. wH. «H. ma mH 5H ma ma qa mm 0H m w! |\ ml (’3 N meom coHuwuawHHo maHQQOSw manomm .>HUGH .mllwv mnsuHso .mllmv ndHonle muuomm .dlle wuduHso .éllov souznollmm Nuummnumm EmuH wocmumHmmm waHUIImH wuoum ummeHUIIwH monm mmuwlmCOIIhH wEHu mwsHm>||mH uHm3 u.anI|mH mnm£u0\3 wmum>COUIIVH 0>HmHomwcHIINH MH®H0\3 wwuw>coolloH m>HmH5mEHIIm uso mummua¢ulm mcHuc5£ :Hmmummllh mCHmSOHMIIm mogm GOmHHmmEOUIIm mnwnu0\3 monmIIN EoUH 199 AUMScHucouv meom COHummHoHuHmm HmHoow .dlllllllllluull ov. «N. «o. no. om. oH. no. oo. oo. mm. mm. HH. om. oo. om. om. mm. oo. mm. nufl>fluow mcHHSU COHflMmexrnHOUIIN. m nm. oo. ma. ma. oo. «o. VH. mo. oH. Hm. oo. Hm. mH. ma. ma. ma. no. oo. muwzu0\3 annomllom HH. oo.| oH. mm. «o. Ho.- NH. mo. oo. oo.- HH. oo. NH. oo.- HH. oo. oo.] macaw manomuumm ma. oo. mo. ma. «o. oo. oo. oo. ma. nH. mo.u oo. oo. ma. oo. no. xcflnauuvm ma. oH.- ma. ma. Ho.a no. mo.u om. ma. Ho. ma. ma. oH. mH.u ma. w>fluoulmm oH. Hm. oo. oo. HH. nH. vo. om. ma. om. oH. oo. ~H.| om. Hw>muauumm oo. oo. om. oo. nH. NH.: No. «a. no. oo.- oH. oH. oH.u ommu ll whdmmemlle mm. ma. mm. ma. oo. ma. HH. HH. ma. oH. Ho. «m. uflmfl> mnwucHllom Hm. nH. ma. mo.- oo.: oo.: oH. oo. oo. mo. «H. uHmH> oconmulov nH. HH. NH.- oo.- oo. oo. Ho. ma. mo. Ho.u mumuumqulov no. mo. mm. oo. HHM om. HH. mmh mm. a: gamelunv oo.- th ma. no. MHM oH. om. oH. “wwucsao>nuo¢ «o. oo. ma. oo. oo. Hm.u Hm. muuomm QDOHm .mllmv mm mm om mm mm Hm om mo ow no ow mo ow mo mo Ho ow om om EwuH 200 no. Nm. no. hH. 0H. NH.1 no. 0H. Ho.l mo.| No. mH. mo.| mo.l mH. mo.l xmmlle oo.! no.1 VH. Ho.| No. 0H.I oo.l NN. mo.l mo.| mo.| No. mo. no. oo.I wsumum 0Ho>omqullmN Ho.| mo.l Mbhl NH.I mH. mo.| Ho. No. No.| No.| wH.I Ho.l mo. No.1 cwEoz UCquOZIIVN mo.l vm. wH.| oo. vo. wo.l mo.l Ho. oH. mo.l No.| mo. oo.l HH.I 0H. th Ho.l No.l mDuMum UHEocooonUOMIIMN oo.| VN. 0N.I No.| mH.l No. no. No. oo.l oH. mN.I mN.I wN. ho.| Nm. mN. HN. wm9 wHQMUIIoH no. 0H.I oo.: no. No.I OH.I oo.l no.1 vo. mo. 50. mo. No.1 mm: MchE ammovmonmlth mo.l mo.| wo.| No.1 wH.I No. Ho.l Ho.l vo. no.1 No. vH.| uHonmmson CH mumu$IIhH wo. vo.| vo. NH.I oo.| No.l mo.| HH. mo. oo.! Ho. AUHHHQOE HMUHm>LmIIwH mH. mo. Ho.| H~.| 0H.I no. mH. MH.I no. wH. :OHummHUHuHmm HMHuomllmH Ho.l oo.l Ho. oo. No. OH. wH.I mH. HH. w uouumm|1VH oo.| No. oo.| vo. mo. mo. mo.l mo. m uOuommIImH Ho. oo.: No.l NH.| mo. no. Ho. v nouommllNH oo.: oo.! Ho. mH.I 0N.I NN.I m uOuommllHH hH.I 0H. mo.| mH.I Ho.l N HeuummIIOH Ho. NH. mo. 8. 3. EGGS—8:0 mchmonmo H nouommllm ho. Ho.l No. no. vo.| >moHoccuwu mcum3ou menuHuUZIIm we. 5H. wH.| Ho. Ho.l v0.1 mcchmn HmcomuwmuoucHlln NH. No. No.1 mo. oo.: mcchmn HMGOmuomEHllw mo.1 3.: 3. 1.- umo E monmuzm oH.| oo.| oH. mCHmmonm HMCOmumeHIIv mH. mN. WEOOEHQUMM \mCch0E moswllm nH. w>MV\mv:0xmw3 mogmilN 33: um. monmlua mN mN VN mN NN HN ON 3 mH nH wH mH VH MH NH .3 0H m m h m m v m N H mmqm BZmQmemQZH ho XHfiHflZ ZOHfiddmmmOU APPENDIX C MULTIPLE REGRESSION STATISTICS APPENDIX C MULTIPLE REGRESSION STATISTICS PROPENSITY TO ENGAGE VARIABLES IN THE USAGE MULTIPLE REGRESSION EQUATION Iggipigignt Simgle Beta F (p) Semipgrtial The Browser .193 .126 3.575(.060) .023 Attitudes Towards Technology* .148 .168 6.712(.010) .022 Number of Cars in Household .189 .104 2.326(.129) .015 The Talker* -.153 - 137 4 477(.o35) .014 Age .110 .110 l.970(.162) .008 Companionship —.082 —.025 2.747(.099) .007 Apathetic .106 .107 2.582(.109) .005 Lifecycle .075 .058 .748(.388) .004 Social Participation —.009 -.062 .781(.378) .003 Sex .039 .028 .150(.699) .002 Socioeconomic Status .028 .061 .815(.367) .002 Education -.010 .044 .381(.538) .001 .060 .038 .298(.586) .001 Working Women 201 202 I32:p:2$:nt Simgle Beta F (p) Semipartial Cable TV .143 .014 .042(.837) .001 Economic —.016 .058 .676(.412) .000 Radio & TV -.005 .025 .l45(.704) .000 Physical Mobility .000 .018 .074(.786) .000 Impulsive .004 -.110 .010(.9l9) .000 Newspapers & Magazines -.017 —.005 .004(.950) .000 * l p Significant Contribution. Key to Tables SHOPPING ORIENTATION VARIABLES Factor O\m.bL0k)H Label The Browser Apathetic Economic Impulsive The Talker Companionship PROPENSITY TO 203 ENGAGE VARIABLES IN THE VALUE MULTIPLE REGRESSION EQUATION Independent Simple Semi artial Variable B Beta F (p) E * Age .278 .284 14.559(.000) .051 , * Lifecycle .276 .199 9.710(.002) .039 Number of Cars in Household .262 .130 4.028(.046) .039 Education* .310 .138 4.193(.o42) .022 Social * Participation -.012 -.159 5.585(.019) .018 Companionship* —.117 —.123 3.822(.025) .007 Economic* .096 .162 5.862(.Ol6) .004 Cable TV .171 .007 .128(.910) .004 The Talker -.058 -.044 .506(.478) .002 Sex .046 .004 .004(.951) .002 Physical Mobility .000 .055 .802(.37l) .001 Impulsive .050 —.017 .072(.789) .001 Apathetic .021 .104 2.704(.10l) .000 The Browser .033 —.069 1.188(.277) .000 Newspapers & Magazines -.013 -.058 .742(.390) .000 Attitudes Towards Technology .001 .025 .l6l(.689) .000 Socioeconomic Status .015 .024 .l43(.706) .000 Radio and Television -.010 .018 .082(.775) .000 Working Women .012 .002 .001(.980) .000 * 0 l u I Significant Contribution. OPPORTUNITY TO 204 ENGAGE VARIABLES IN THE USAGE MULTIPLE REGRESSION EQUATION Independent Simple Semipartial Variable B Beta F (p) Interpersonal Shopping .202 .143 3.350(.069) .023 Shop at Night .183 .149 3.479(.064) .019 Impersonal Shopping .102 .090 1.366(.244) .013 Shop in the Afternoon/ Morning .048 .002 .000(.984) .004 Interpersonal Banking —.055 -.020 .069(.794) .002 Shop on Week- ends/days -.041 —.039 .247(.620) .001 Impersonal Banking .014 .032 .179(.673) .000 205 OPPORTUNITY TO ENGAGE VARIABLES IN THE VALUE MULTIPLE REGRESSION EQUATION Independent Simple Semip rtial Variable B Beta F (p) r31 Impersonal Banking .085 .102 1.776(.l84) .011 Shop at Night .078 .071 .752(.387) .005 Shop on Week- ends/days —.050 -.073 .856(.356) .004 Interpersonal Shopping —.057 —.O60 .573(.450) .003 Shop in the Afternoon/ Morning .058 .052 .409(.523) .002 Interpersonal Banking —.327 —.049 .393(.532) .002 Impersonal Shopping .027 .013 .029(.865) .000 206 PROPENSITY AND OPPORTUNITY TO ENGAGE VARIABLES IN THE USAGE MULTIPLE REGRESSION EQUATION Independent Sim 1e ' ' Variable BP Beta F (p) Sem1p3rt1al Shop at Night .183 .109 2.389(.124) .028 Attitudes Towards Technology .148 .158 6.092(.014) .020 Interpersonal Shopping .202 .181 6.869(.009) .018 The Talker* —.153 —.145 5.148(.024) .016 Age* .110 .167 4.138(.043) .016 The Browser .193 .104 2.329(.128) .014 Number of Cars in Household .189 .084 1.515(.220) .011 Impersonal Shopping .102 .099 2.274(.133) .010 Companionship —.082 -.127 3.525(.062) .009 Apathetic .106 .059 .750(.387) .007 Lifecycle .075 .050 .559(.455) .003 Shop on Week- ends/days -.O41 -.059 .757(.385) .002 Social Participation —.009 —.042 .345(.558) .002 Interpersonal Banking -.055 -.032 .23l(.63l) .002 Sex .039 .030 .176(.675) .002 Socioeconomic Status .028 .078 l.326(.251) .001 Economic -.016 .056 .630(.428) .001 207 Independent Simple Semipartial Variable B Beta F (p) r Working Women .060 .050 .528(.468) .001 Radio & Tele— vision -.005 .033 .27l(.603) .001 Newspapers & Magazines —.017 —.012 .030(.862) .001 Cable TV .143 .008 .016(.899) .001 Shop in the Afternoon/ Morning .048 —.056 .580(.447) .000 Physical Mobility .000 .036 .308(.580) .000 Impersonal Banking .014 .025 .153(.696) .000 Education .010 .023 .108(.743) .000 Impulsive .004 -.002 .001(.972) .000 * I I Significant Contribution. 208 PROPENSITY AND OPPORTUNITY TO ENGAGE VARIABLES IN THE VALUE MULTIPLE REGRESSION EQUATION Inde endent Sim 1e » . VarIable Bp Beta F (p) Semipartial * Age .278 .295 13.730(.000) .048 * Lifecycle .276 .201 9.708(.002) .039 Number of Cars in Household .262 .109 2.734(.100) .029 Education* .310 .164 5.750(.017) .025 Social Participation -.012 -.128 3.412(.066) .014 * Economic .096 .156 5.167(.024) .012 Companionship* -.117 -.142 4.682(.032) .012 Impersonal Banking" —.085 .130 4.224(.o41) .011 Interpersonal Banking —.327 —.063 .943(.333) .006 Shop at Night .078 .054 .630(.428) .005 Shop on Week- ends/days —.050 -.083 1.577(.210) .004 The Talker —.057 -.054 .749(.388) .003 Sex .046 .022 .099(.753) .003 Physical Mobility .000 .065 l.077(.300) .002 Interpersonal Shopping —.057 —.023 .ll9(.731) .002 Im ersonal Shgpping .027 .032 .257(.613) .001 Impulsive .050 -.002 .001(.976) .001 .171 .000 .000(.999) .001 Cable TV 209 Independent Simple Semipartial Variable B Beta F (p) r Apathetic .021 .065 .97l(.325) .000 The Browser .033 -.063 .970(.343) .000 Newspapers & Magazines -.013 -.043 .402(.526) .000 Shopping in the Afternoon/ Morning .058 —.031 .186(.667) .000 Socioeconomic Status .015 .024 .l37(.712) .000 Radio & Television .010 .023 .136(.713) .000 Working Women .012 .017 .068(.795) .000 Attitudes Towards Technology .001 —.014 .049(.825) .000 * I I Significant Contribution. SELECTED BIBLIOGRAPHY SELECTED BIBLIOGRAPHY "A Retreat from the Cashless Society. " Business Week, April 18, 1977, p. 81. "Academy Study Finds Low Energy Growth Won' t Be Painful. Science 195 (January 28,1977): 380. Albertson, Lesley A. "Telecommunications as a Travel Substitute: Some Psychological, Organizational and Social Aspects." Journal of Communication (Spring 1977): 33. ~ Anderson, Nels, and Lindeman, Edward C. Urban Sociology. New York: Alfred A. Knopf, Inc., 1928. "ATMS vs. Passbooks." American Banker, October 13, 1976, p. 4. "Autotransaction Industry Report." Autotransaction, March 3, 1976, p. 4. Baer, Walter S. Cable Television: A Handbook for Decisionmaking. New York: Crane, Russak & Company, Inc., 1974. Bagdikian, Ben H. The Information Machines. New York: Harper and Row, 1971. Blalock, Hubert M., Jr.‘ Social Statistics. New York: McGraw—Hill Book Company, 1972. Block, Martin P.; Wright, James B.; and McVoy, D. Stevens. "An Evolutionary Approach to the Development of Two- Way Cable Technology Communication. IEEE Transactions on Communications (publication forth- coming). Branscomb, Anne W. "The Cable Fable: Will It Come True?" Journal of Communication (Winter 1975): 50—62. 210 211 Human Activity Patterns in the New Chapin, F. Stuart, Jr. City, Things People Do in Time and Space. York: John Wiley & Sons, 1974. Retail Management-—A Channels Calif.: Wadsworth Publish- Dickinson, Roger A. Ap roach. Belmont, ing Company, Inc., Ford Foundation Study Urges Action on Conser— Science 186 (November 1, 1974): 426. "Energy: vation." "The WSFS Plan—-A Practical Experiment The Magazine of Bank Administration, 1974. Evers, Francis P. in EFTS." April, 1976, pp. 52—55. "Communication Substitutes for The Movement of Information Paper presented Hanneman, Gerhard J. Transportation: Versus the Movement of People." to International Communication Association, New Louisiana, April 1974. Richard C. Telecommunications Substitutes for Travel. Special Report to the Office of Tele- U.S. Department of Commerce: December 1973. communications, OT—SP—73—2, COM-74—10075, and Sheth, Jagdish N. The Theory of New York: John Wiley and Sons, Orleans, Harkness, Howard, John A., Buyer Behavior. 1969. COM-23, No. 10 Inc., IEEE Transactions on Communications, (October 1975). Jones, Douglas, ed. Communication and Energy in Changing Urban Environments. London: Butterworth and Co., Ltd., 1970. "The Impact of Communications Technology: Journal of Communication Katzman, Natan. Promises and Prospects." 47—58. Elazar J. Multiple New York: 1973. (Autumn 1974): Kerlinger, Fred M., and Pedhazur, Regression in Behavioral Research. Holt, Rinehart and Winston, Inc., "Columbus Viewers Will Take Part in Pro— Koris, Sally. grams When Two—Way Service Begins Late in the The wall Street Journal, August 1977, Year." p. 32. 212 LaPorte, R., and Metlay, Daniel. "Technology Observed: Attitudes of a Wary Public." Science (April 17, 1975): 121-27. Prentice- Larson, Carl M.; Weigand, Robert B.; and Wright, John S. Englewood Cliffs, N.J.: Basic Retailing. 1976. Telecommunications Substitutability Inc., An Energy Conservation Potential. Hall, Lathey, Charles E. for Travel : Report to the Office of Telecommunications, U.S. Department of Commerce, OT Report 75-58, January 1975. Lavidge, R., and Steiner, G. A. "A Model for Predictive Measurements of Advertising Effectiveness." Journal of Marketing 25 (1961): 59—62. "Cable TV in Columbus: Ledingham, John, and Henning, Todd. Can It Deliver?" Columbus Monthly, June 1977, pp. 46-51. Massy, William F.; Montgomery, David B.; and Morrison, Donald G. Stochastic Models of BuyingiBehavior. Cambridge, Mass.' The MIT Press, 1970. Nicosia, Francesco. Consumer Decision Processes. Englewood Cliffs, N.J.: Prentice-Hall, 1966. Jenkins, eds. Nie, N. J.; Hull, C. H.; J. G.; Steinbrenner, K.; and Brent, D. H., Statistical Package for the Social Sciences. New York: McGraw Hill, 1975. Carlson, F. R., Jr.; Gray, P.; and Hanneman, The Telecommunications—Transportation Options for Tomorrow. New York: 1976. J. M.; G. J. Tradeoff: John Wiley and Sons, Study Claims Effects Would Be Modest, Nilles, "Nuclear Moratorium: Foresees Low Growth Rate for Total Energy Demand." 1977): 156—57. University Science 195 (January 14, Minneapolis: They went to Coféjge. Pace, C. R. of Minnesota Press, "Planning and Experience in Automated Teller Machines." June 15, 1976, pp. 23-25. P001, Ithiel de Sola, ed. Talking Back: Citizen Feedback and Cable Technology. Cambridge: The MIT Press, 1973. Bank News, 213 Preston, Lee E., ed. New Research in Marketing. Berkeley: University of California Institute of Business and Economic Research, 1965. Redinbaugh, Larry D. Retailing Management—-A Planning Approach. New York: McGraw Hill Book Company, 1976. Reid, Alex. New Directions in Telecommunications Research. A Report to the Sloan Commission on Cable Communi- cations, 1971. Rich, Clyde L. "Is Random Digit Dialing Really Necessary?" Journal of Marketing Research 14 (August 1977): 305. Rich, Stuart U. Shopping Behavior of Department Store Customers. Boston: Harvard University Graduate School of Business Administration, 1963. Richley, Robert D. "Branch Banks and CBCTS: A Delicate Balance." Bankers Monthly Magazine, March 1976. Rogers, Everett, and Shoemaker, F. Floyd. Communication of Innovations. New York: The Free Press, 1971. Schnore, Leo F., ed. Social Science and the City: A Survey of Urban Research. New York: Frederick A. Praeger, 1968. Scott, James D. Cable Television, Strategy for Penetrat- ing Key Urban Markets. Michigan Business Report No. 58. Ann Arbor: University of Michigan Press, 1976. Sloan Commission. On the Cable—-The Television of Abun— dance. New York: McGraw—Hill Book Company, 1971. Sonquist, John A. Multivariate Model Building, Validation of a Search Strategy. Ann Arbor, Mich.: Insti— tute for Social Research, University of Michigan, 1970. Stone, G. "City Shoppers and Urban Identification: Observations on the Social Psychology of City Life." The American Journal of Sociology 60 (July 1954): 37-45. Szalai, Alexander, ed. The Use of Time-Daily Activities of Urban and Suburban Populations in Twelve Countries. The Hague: Mouton and Company, 1972. 214 Tauber, Edward M. "Why Do People Shop?" Journal of Marketing 36 (October 1972): 48. Technology as a Deterrent to Dehumanization." Science 185 (August 30, 1974): l. \lHlNHIHHIN"WNWIIHHIHIHUHIHHIWIWWI