'_ a a u - v. 'v..»I" ' ' IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 3 1293 10499 11 h“"““i~~¢r~ W-. 3. . ,. m h a , Mlyhh.‘. ”End This is to certify that the dissertation entitled Physical Distribution Service - A Comparative Study presented by Fernando Bins Luce has been accepted towards fulfillment of the requirements for Ph.D. degreein Business Administration Major pro essor Date March 11, 1982 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 MSU LIBRARIES ‘32—‘- RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. PHYSICAL DISTRIBUTION SERVICE: A COMPARATIVE STUDY By Fernando Bins Luce A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Marketing and Transportation Administration 1982 @ Copyright by FERNANDO BINS LUCE 1982 ABSTRACT PHYSICAL DISTRIBUTION SERVICE: A COMPARATIVE STUDY By Fernando Bins Luce This dissertation presents a report of research about physical distribution service in Brazil. The major thrust of this study was to make a comparative analysis of the findings of this research with those - reported by Perreault and Russ (1976a) in the United States. Also, the investigator sought to examine relationships not reported in the Perreault and Russ study so as to provide additional understanding of physical distribution service in the Brazilian environment. The population of the study comprised purchasing managers of companies with more than 50 employees, located in either the metro- politan area of Porto Alegre or in the city of Caxias do Sul, both in the State of Rio Grande do Sul, and within a certain group of indus- tries (metallurgical, mechanical, electrical, transportation equipment, furniture, plastic, and shoe). Because it was feasible and desirable, all firms within the population were invited to participate in the study; thus, no sampling was necessary. The data were gathered by a questionnaire in which the respond- ents were asked to answer questions on product-specific situations. The questionnaire was administered to 418 subjects. The return rate Fernando Bins Luce was 43% (l80 questionnaires). To minimize nonrespondent error, a sample was drawn from this population, and personal interviews were conducted with 24 nonrespondents. Then tests for differences in responses were performed. No relevant differences were found. The major variables of the study were divided into two groups. The independent variables included supplier (type, number), company (size, industry category), and situational variables (deliveries, backorders, order cycle time). The dependent variables were satisfac- tion (physical distribution service and its components), importance (physical distribution service components), and purchasing factors variables. The major differences between the two studies centered on the following aspects: (l) the importance of price and physical distribu- tion service as factors in selecting suppliers and (2) the relation- ships between number of deliveries and satisfaction with physical distribution service with its importance in selecting suppliers. The relevant similarities included the identical pattern concerning satis- faction with service and feedback about service needs, the highest importance ranking of product quality as a factor in selecting sup- pliers, and the unobserved relationships between number of suppliers and backorders with importance of physical distribution service in selecting suppliers. ACKNOWLEDGMENTS Dr. Donald Taylor, chairman of my committee, provided the necessary guidance throughout my doctoral program. His knowledge and competence in the field of marketing and physical distribution enriched the content of this dissertation, and his encouragement and availa- bility were instrumental in the culmination of this report. The other committee members, Drs. George Nagenheim and David Closs, offered suggestions that improved this dissertation. Ms. Gabriella Belli of the Office of Research Consultation gave competent and professional assistance in the statistical analy- sis of the data. My research assistants in Porto Alegre provided the necessary help in the data collection and preparation. Maria Nina Braga par- ticipated in all stages of that process. Fani Ferlauto helped in the prenotification phase, and Luis Antonio Slongo participated in the tests for nonrespondent error and in the data preparation. My colleagues at the Universidade Federal do Rio Grande do Sul were very supportive in this endeavor. Joao Luis Becker assisted in the questionnaire design, in the preparation and coding of the data, and in the selection of statistical techniques. Dr. Dennis Alan Guthery suggested the theme of this dissertation. ii Dr. Roberto Costa Fachin, coordinator of the Graduate Program of Administration at the Universidade Federal do Rio Grande do Sul, gave his support and encouragement for this dissertation. Edi Fracasso, chairperson of the Department of Administrative Sciences, and Antonio Carlos Rosa, dean of the School of Economics, both at the Universidade Federal do Rio Grande do Sul, freed me of all academic and administrative duties to allow for the completion of this dissertation. CAPES offered the necessary financial support in all my doc- toral studies and in the dissertation. FINEP, FAPERGS, and the Uni- versidade Federal do Rio Grande do Sul funded the data collection. Susan Cooley provided editorial assistance and typed the final version of this dissertation. Biti, my companion in all this endeavor, offered competent and timely suggestions. Her understanding of the value and meaning of a doctoral degree were the basis of her continuous support and encouragement. TABLE OF CONTENTS OK!) to oooowwuosmwN—I Page LIST OF TABLES ......................... vii LIST OF FIGURES ......................... X Chapter I. INTRODUCTION ...................... l Purpose of the Study ................. Importance of the Study ................ Conceptual Framework ................. Research Objectives .................. Hypotheses ...................... Limitations ...................... The Instrument ................... The Population ................... The Analyses .................... Overview of the Dissertation ............. II. LITERATURE SEARCH .................... Introduction ..................... Customer Service ................... Components of Customer Service ........... 12 Physical Distribution Service ............. 13 Definition of P03 .................. 15 P05 Elements .................... 16 Research in the Area of P05 .............. l7 PDS Components ................... l7 P05 and the Selection of Suppliers ......... 22 Factors Affecting PDS ................. 26 Framework of This Research .............. 27 III. RESEARCH METHODOLOGY .................. 30 Population ...................... 30 Response Problems ................... 32 Sponsorship of the Research ............. 32 Prenotification ................... 32 Data-Collection Instrument .............. 33 iv Types of Products .................. Test of the Instrument ............... Data-Collection Procedures .............. Research Objectives and Methodology .......... Objective 1: Satisfaction .............. Objective 2: Importance ............... Objective 3: Importance ............... Objective 4: Relationships ............. Objective 5: Comparative Analysis .......... Other Tests ..................... IV. RESEARCH FINDINGS .................... General Descriptive Findings ............. Respondents ..................... Suppliers ...................... Purchasing Situation ................ The Chosen Product ................. Other General Descriptive Findings ......... Satisfaction With PDS and Its Components ....... Descriptive Findings ................ Relational Findings ................. Importance of PDS Components ............. Descriptive Findings ................ Relational Findings ................. PDS and Other Factors Affecting the Patronage Decision . Descriptive Findings ................ Relational Findings ................. PDS and the Patronage Decision ............ Situational Variables ................ Supplier Variables ................. Company Variables .................. Satisfaction Variables ............... Comparative Analysis ................. Satisfaction With PDS and Its Components ...... Importance of Factors in Selecting Suppliers . . Relationships .................... Summary of Findings .................. V. CONCLUSIONS ....................... Satisfaction Nith PDS ................. Importance of PDS Components ............. Factors Affecting the Patronage Decision ....... PDS and the Patronage Decision ............ Comparative Analysis ................. 105 109 109 Satisfaction Nith PDS and Its Components ...... 119 Importance of Factors in Selecting Suppliers . . . . 119 Relationships .................... 120 Suggested Areas of Further Research .......... 121 APPENDICES ........................... 123 A. CUSTOMER SERVICE ELEMENTS ................ 124 B. QUESTIONNAIRE ...................... 131 C. LIST OF VARIABLES .................... 148 BIBLIOGRAPHY .......................... 152 Vi LIST OF TABLES Table 2.1. Rankings of Selected Factors by Product Type ....... 3.1. The Study Population, by Industry and Region ....... 3.2. Questionnaire Responses, by Industry and Location 4.1. Activities of the Respondents .............. 4.2. Distribution of "Other" Activities of the Respondents 4.3. Distribution of Respondents by Company Size (Number of Employees) ....................... 4.4. Distribution of Respondents by Company Size (Sales Volume) ........................ 4.5. Distribution of Respondents by Type of Industry ..... 4.6. Distribution of Number of Suppliers Utilized ....... 4.7. Distribution of Number of Other Suppliers Available 4.8. Distribution of Number of Alternative Suppliers ..... 4.9. Supplier Categories ................... 4.10. Distribution of Number of Deliveries in 1980 ....... 4.11. Distribution of Percentage of Backorders ......... 4.12. Distribution of the Average Order Cycle Time ....... 4.13. Distribution of the Chosen Products ........... 4.14. Changes of Suppliers in the Last Two Years ........ 4.15. Distribution of Major Reasons for Changing Suppliers . . 4.16. Satisfaction Ratings for PDS Components ......... 4.17. Summary of the Wilcoxon Test for Mean Ratings of Satisfaction for PDS Components ............ vii Page 23 31 37 52 52 53- 54 54 55 56 56 57 58 58 59 6O 63 63 64 65 .18. .19. .20. .21. .22. .23. .24. .25. .26. .27. .28. .29. .30. .31. .32. .33. .34. .35. Distribution of Satisfaction Ratings for the Overall PDS .......................... Summary of t-tests With Supplier Category ........ Correlations Between Supplier Variables and Overall PDS . MANOVA: Satisfaction With PDS Components by Company Size (Employees) ................... ANOVA: Satisfaction With Overall PDS (Question 23) by Company Size (Employees) ............... Correlations Between Situational Variables and Overall PDS .......................... Summary of t—tests With Variable Percentage Backorders Importance Ratings for PDS Components .......... Summary of the Wilcoxon Test for Mean Ratings of Importance for PDS Components ............. Summary of t-tests With Supplier Category ........ Correlations Between Company Variables and Importance of P05 Components ................... Summary of ANOVAs Between Company Size and Importance of PDS Components ................... Summary of the Mean Rankings of Importance for Different Groups of Company Size (in Terms of Number of Employees) ..................... Summary of ANOVAs Between Company Size and Importance of PDS Components ................... Summary of the Mean Rankings of Importance for Different Company-Size Groups (in Terms of Sales Volume) MANOVA: Importance of PDS Components by Company Size (Employees) ...................... MANOVA: Importance of PDS Components by Company Size (Sales) ........................ Importance Ratings of Purchasing Factors ........ viii Page 67 68 69 71 71 72 73 74 ’ 75 76 78 79 79 80 81 82 83 84 4.38. 4.42. 4.43. 4.46. 4.47. 4.48. 5.1. 5.2. Summary of the Wilcoxon Test for Mean Ratings of the Importance of the Purchasing Factors ........ Summary of t-tests With Supplier Category ....... Correlations Between Company Variables and Factors in Selecting Suppliers ................. MANOVA: Importance of Purchasing Factors by Company Size (Employees) .................. MANOVA: Importance of Purchasing Factors by Company Size (Sales) .................... Breakdown of Average Order Cycle Time by Importance of PDS ....................... Spearman Correlation Coefficients ........... Summary of ANOVAs Between Company Size and Importance of PDS ....................... Mean Rankings of Importance of PDS by Company Size (Number of Employees) ................ Mean Rankings of Importance of PDS by Company Size (Sales Volume) ................... Satisfaction Ratings of PDS Components ........ Correlations Between Components and Overall PDS . . . . Supplier Sensitivity to Purchasers' Service Needs and Purchaser Satisfaction With PDS ........... Ranks of Importance of Purchasing Factors ....... Mean Responses on Variables by Importance of PDS Groups of Satisfaction Ratings and Correlations . . . . Mean Importance Rankings of PDS Components ...... ix Page 85 86 87 88 89 91 92 94 94 95 97 98 99 100 103 113 114 LIST OF FIGURES Figure Page 2.1 Suppliers Seeking Feedback About Physical Distribution Services ....................... 61 2.2 Actions Concerning a Stockout Possibility ........ 62 CHAPTER I INTRODUCTION This dissertation reports research that was conducted to examine different aspects of physical distribution service in the Brazilian business environment and to compare them with a study done by Perreault and Russ (1976a) in the United States. This chapter includes an overview of the purpose and the importance of the study, followed by a brief analysis of the conceptual framework of the research. The research objectives are introduced, and the major limi- tations of the methodological procedures are examined. Purpose of the Study This study was conducted to enable further acquisition of knowledge in the area of physical distribution by reporting a compara- tive analysis of perceptions of purchasers concerning the physical distribution service provided by their suppliers in two different business environments: Brazil and the United States. Physical distribution service was selected for study because it is the overall objective of a physical distribution system. The Perreault and Russ study was selected for comparative purposes because it lends itself to replication in another environment. Repeating a similar trend that has occurred in the United States in the last three decades (Bowersox, 1978, pp. 4-12), the 1 concept of physical distribution is evolving in Brazil. In conduct- ing this research, the investigator was concerned with the generali- zation and application, to the Brazilian situation, of the findings reported by Perreault and Russ (1976a) about physical distribution service in the United States because there are considerable dif- ferences between the two countries. There is a substantial gap in the countries' economic development, which is certainly reflected in their business practices. The higher inflation rates in Brazil have a direct effect on interest rates, which in turn affects the supply and demand of credit that influences the management of funds within a single business enterprise. The not-so-legally pro- tective competitive system in Brazil permits the development of monopolistic and oligopolistic market structures that affect the busi- ness environment. All of the aforementioned environmental character- istics directly influence physical distribution service from the point of view of both suppliers and customers. The investigator also sought to examine relationships not reported in the Perreault and Russ study so as to provide additional understanding of physical distribution service in the Brazilian envi- ronment. This was done with the intention of augmenting the limited body of knowledge about business in Brazil, in the hope of contribut- ing to the development of a unique theoretical framework to deal with specific problems of physical distribution service in Brazil. Importance of the Study The Brazilian economy is experiencing a host of economic prob- lems that have a direct effect on the area of physical distribution. These problems include high interest rates, inflation, high costs of energy, and tight monetary policies. The interest rates and the high inflation pose extreme restrictions on inventory policies concerning availability, which in turn influences the level of physical distri- bution service that can be offered to customers. The high costs of energy are posing threats to the stability of Brazil's transportation system. Physical distribution service levels will certainly have to be revised. Also, the tight monetary policies are bringing about a reduction in consumer demand, resulting in a profit squeeze that reduces the funds available for providing adequate physical distri- bution service. At the same time, the knowledge of physical distri- bution is growing within the Brazilian business environment. Therefore, the need for insights regarding physical distribution service practices and perceptions arises. The findings of this research can foster an improvement in business practices in Brazil by providing elements to further managerial competence through new insights about physical distribution service as perceived by the recipients of such services. Conceptual Framework In spite of extensive research in the area of physical dis- tribution service, it seems that the authors have not yet reached consensus about what is meant by physical distribution service and what constitutes its components. The problem may have originated in the confusion that exists in differentiating customer service from physical distribution service. The former should be viewed as a set of activities or elements that constitute the interface between a company and its customers, whereas the latter is one of the elements of that interface--that is, physical distribution. Therefore, in this dissertation physical distribution service was viewed as one ele- ment of customer service, which was defined as "those activities that occur at the interface between the customer and the corporation which enhance or facilitate the sale and use of the corporation's products and services" (La Londe & Zinszer, 1976, p. 2). Definitions of physical distribution service range from one extreme, which emphasizes the time-and-place utilities provided, to another, in which emphasis is placed on either highlighting the com- ponents or elements of physical distribution service or establishing performance standards to be attained by physical distribution service. For the purpose of this research, physical distribution ser- vice was defined as "the interrelated package of activities provided by a supplier which creates utility of time and place for a buyer and increases form utility" (Perreault & Russ, 1976a, p. 3). This defini- tion was operationalized by a number of components that constitute the physical distribution service package, which is the same as the one described by Perreault and Russ (1976a, p. 8). Therefore, in this dissertation the components of physical distribution service included billing procedures, average delivery time, delivery time variability, rush service, returns policy, order status information, accuracy in filling orders, action on complaints, and order methods. Throughout the remainder of the dissertation, the abbreviation PDS is used and refers to physical distribution service as defined above. The basic framework of this research stemmed from the work done by Perreault and Russ (1976a) so as to provide the means for a comparative analysis. Therefore, the categories of variables and some of the hypotheses are similar in both studies. Additional variables and hypotheses were introduced in the present work to allow for the fulfillment of the other major purpose of this research. The cate- gories of variables are outlined at the end of Chapter II, and the hypotheses are presented later in this chapter. Research Objectives The operationalization of this research undertaking was accomplished by a certain number of objectives and hypotheses. The first objective concerned satisfaction with the overall PDS and its components. The second was intended to rank the PDS components in order of importance. The third objective sought to compare PDS with other factors in patronage decisions. The fourth explored relation- ships of particular variables to the importance of PDS in selecting suppliers. For this objective, a series of six substantive hypothe- ses were stated. The fifth objective was to compare the findings of this study with those presented by Perreault and Russ (1976a). The specific research objectives were as follows: 1. To investigate the perceived satisfaction of buyers with overall PDS and with each of its components. 2. To rank, in order of importance, different components of PDS as perceived by purchasers. 3. To compare the importance of PDS with other factors influ- encing patronage decisions. 4. To explore relationships between situational variables of the buying process, supplier variables, company variables, and satisfaction variables with the perceived importance of PDS in selecting suppliers. 5. To present a comparative analysis of the findings of this study with those reported by Perreault and Russ (1976a). This analy— sis compares the findingsirlthe Brazilian and American environments on the following aspects: (a) satisfaction with overall PDS and each of its components, (b) the importance of PDS as a factor influ- encing patronage decisions, and (c) relationships involving the per- ceived importance of P05 in selecting suppliers. Hypotheses The hypotheses derived from the fourth objective are presented below: 1. The greater the number of deliveries, the greater the importance of PDS in selecting suppliers. 2. The higher the proportion of backorders, the greater the importance of PDS in selecting suppliers. 3. The greater the average order cycle time, the greater the importance of PDS in selecting suppliers. 4. The greater the number of alternative suppliers avail- able, the lower the importance of PDS in selecting suppliers. 5. The larger the company, either in terms of number of employees or in sales volume, the greater the importance of PDS as a factor in selecting suppliers. 6. The greater the satisfaction with PDS, the lower its importance as a purchasing factor. Limitations The limitations of this research stemmed from three basic sources: the instrument, the population, and the analyses. Each is discussed below. The Instrument The data-collection instrument was relatively long and fairly complicated. Most of the questions were product and supplier specific, and some of them related to different time frames. The responses were based on self-support, with the exception of a sample of nonrespond- ents who were personally contacted. A number of questions required perceived responses, as opposed to specific data-based answers. The products chosen for the questionnaire were all standardized industrial products with a sizable number of alternative suppliers. The concep- tual background used in developing the instrument was, in itself, a limitation. Since the Perreault and Russ (1976a) study was used as a starting point for this research undertaking, the present effort was in a sense limited to their framework. The Population The criteria used to define the population (see Chapter III for details) were fairly restrictive. First of all, a defined geo- graphical area prevented the inclusion of a large number of indus- tries in the population. The available roster of members of the Federacao das Indfistrias do Rio Grande do Sul, from which the respond- ents were selected, had data on the companies for the year 1979; some subjects might have been excluded from the population because of out- dated information in the roster. And finally, only a selected group of industries (see Table 3.1) was included in the population. The Analyses Most of the analyses were either descriptive or relational and were based heavily on a correlational model. Therefore, causality may not be inferred from this type of analysis. Overview of the Dissertation The core of this research undertaking is presented in the next four chapters. In Chapter II the literature on customer service is examined, with emphasis on PDS. A discussion of the various definitions of customer service and PDS is presented, as well as a review of the recent research findings in the area of PDS. In Chapter III the methodology developed for this research is discussed. The population of the study is described, and the data-collection instrument and procedures are presented. The statistical tools used for testing the hypotheses and exploring relationships are introduced. Chapter IV contains a summary of the substantial findings of the study, and Chapter V outlines the major conclusions and implications of this investigation and explores further research that could be derived from this endeavor. CHAPTER II LITERATURE SEARCH Introduction This study on the role of physical distribution service (PDS) explores buyers' perceptions of satisfaction with and impor- tance of certain PDS components and the importance of PDS as a factor in selecting suppliers. The study stems from previous work done by Perreault (1973) and by Perreault and Russ (1974, 1976a) but includes a broader theoretical background that examines more recent research developments in the field of PDS. In this chapter, the investigator reviews major concepts and research approaches and findings that provide a background for the study. The literature examined in this chapter directly influenced the selection of research variables, the design of the data-collection instrument, and the final interpretations of the study's findings, conclusions, and recommendations. Customer Service Since Perreault's investigation, a host of different research studies have addressed the area of customer service, with emphasis on the demand-obtaining aspect of customer service. In spite of the exten- sive work in this area, it seems that some of the original difficulties in determining what is meant by customer service still persist. In this 10 respect, Daniel and Jones (1969) emphasized that "anyone who has attempted to struggle with a definition of customer service realizes that it is a paradoxical concept" (p. 344). The major problems appear to be in defining the scope of customer service and also in assigning responsibilities for managing customer service within the organiza— tional structure. La Londe and Zinszer (1976) summarized this situa- tion by highlighting three different corporate approaches to dealing with customer service. The first considers customer service as a set of activities within the firm. Examples are organizing and implement- ing billing procedures, customer complaints, and so on. The second approach views customer service in terms of performance levels, which could be related to availability of product, returns due to damages in transit, order cycle time, etc. The third approach regards cus- tomer service as "an element of total corporate philosophy" (Rose, 1979, p. 280). This approach considers customer service as a new area of managerial concern and could be viewed as similar to the idea of the marketing concept. Regardless of the approach one uses to view customer service, it is undoubtedly an area that permeates almost all of the managerial functions within an organization. Definitions of customer service range from the most diffuse and general to statements that comprise specific indicators of customer service performance. Among the broader definitions, it is appropriate to mention the following: Customer service is a chain of events that is in the business of keeping customers. (Davis, 1971, p. 51) 11 Customer service constitutes those activities that occur at the interface between the customer and the corporation which enhance or facilitate the sale and use of the corporation's products and services. (La Londe & Zinszer, 1976, p. 2) Customer service is a complete collection of demand-related factors under the control of the firm, but whose importance in determining supplier patronage is ultimately evaluated by the customer receiving the service. (Ballou, 1973, p. 96) Customer service is a function which is concerned with all of the operating interfaces, except selling, between the company and its customers. (NCPDM, 1978, p. 188) Toward more specific definitions of customer service, one may consider those that state the economic utilities fulfilled by customer service. They are: Technical product services are those activities which the manu- facturer engages in besides sale of the product, to produce for the purchaser of his product the expected utility or end values in terms of product performance for which the product is obtained. (Simon, 1965, p. 32) [This definition deals with form utility.] Customer service is the quality with which the flow of goods and services is managed. (Ballou, 1978, p. 62) [Temporal and spatial utilities are implicit in this definition.] Customer service is the interrelated package of activities pro- vided by a supplier which creates utility of time and place for a buyer, and insures form utility. (Perreault, 1973, p. 15) These last two definitions address the concept of physical distribution service (PDS), which is the major topic of the present study. The literature has not developed a definition of customer service that deals only with possession utility. However, several authors have included dimensions that are related solely to possession utility as being part of either product service or physical distribution service. Examples are visitation services (Simon, 1965, p. 33), terms of sale and ease of ordering (Levy, 1981, p. 91), and salesmen's visits (Perreault & Russ, 1974, p. 40; Gilmour, 1979, p. 87). There 12 is room for the development of a customer—service definition that would consider only possession utility. Components of Customer Service La Londe and Zinszer (1976, pp. 272-82) presented the most comprehensive description of the customer service components, which grouped the elements of customer service into three distinct cate- gories: the pretransaction elements, which provide the organizational posture for customer service; the transaction elements, which fulfill the delivery function; and the posttransaction elements, which support the product while in use. Perreault and Russ (1974, p. 40), Cunningham and Roberts (1974, p. 22), and Gilmour (1979, p. 87) also provided lists of customer service components. Cunningham and Hardy (n.d., p. 125) conducted research on the components of service. They investigated the elements of service as perceived by retailers in experimental studies in Great Britain. The findings showed the following elements of service to be important: reliability of delivery, call frequency, delivery frequency, availa- bility of stock, personal relationship with salesman, and provision of advice and information. The findings were situation specific, and the authors did not mention the reasons for including the elements listed above as part of customer service. La Londe and Zinszer (1976, pp. 23-24) conducted an extensive research project sponsored by the NCPDM to gather information on the elements of customer service. The categories of service were the 13 following: product availability, order cycle time, distribution- system flexibility, distribution-system information, distribution- system malfunctions, and postsale product support. The categories were obtained by personal interviews with physical distribution executives in U.S. industries. A second phase of the research was a survey that centered on the importance of the customer service ele- ments, costs, definition of customer service, marketing variables, and other situational questions. Even though the title of the study was Customer Service, it is evident that the research aimed at PDS. Again, the problem of correctly defining customer service persisted throughout the study. The definition provided in the introduction to the research report was very broad and encompassed the activities involved in the customer-corporation interface for the purpose of facilitating and enhancing the sale and use of a product. After examining the findings of the research, it seems possible that the respondents had a different perception concerning the definitions of customer service (see pp. 203-17). At the end of the study, the authors acknowledged this possibility. Physical Distribution Service In the last decade, the field of logistics has become estab- lished as an area of managerial concern. The climate of the times, with the shortages of raw materials and of reliable energy sources and the economic problems of increasing inflation rates and rising unemployment, has paved the way for the growing importance of logis- tics in the management of organizations. 14 According to Bowersox (1978), logistics is "the process of strategically managing the movement and storage of materials, parts and finished inventory from suppliers, between enterprise facilities and to customers" (p. 3). From this definition, it is clear that logistics deals with movement and storage of goods; hence the per- formance of logistical operations results in fulfillment of time- and-place utilities. One of the areas of logistics is physical distribution, which "deals with the movement, storage and processing of orders for a firm's output" (Ballou, 1978, p. 26). In providing time-and-place utilities, physical distribution is confronted by two conflicting objectives: (1) to maximize customer service (i.e., time-and-place utilities) and (2) to do so at the least total cost of offering the service. Management is dealing with a trade-off situation because “no physical distribution system can simul- taneously maximize customer service and minimize distribution cost" (Kotler, 1967, p. 420). The idea of maximizing customer service is subordinated to the posssibility of establishing a demand-obtaining function related to customer service. Since customer service involves a host of dif- ferent components, the development of demand functions has proven to be cumbersome. Thus, until the early 19705, the majority of the writ- ings and research concentrated on the minimization problem of physical distribution (Perreault, 1973, pp. 13-15), i.e., how to provide a cer- tain level of service at the lowest cost. Nevertheless, scholars and researchers have acknowledged the effect of customer service on 15 generating sales (Hutchinson & Stolle, 1968; Johnson & Parker, 1961; Magee, 1960; Stephenson & Willett, 1968; Stolle, 1967). Recently, the increasing use of quantitative techniques, with the aid of improved computer technology in management investigation, has allowed researchers to shift emphasis toward the demand-obtaining aspect of customer service. Before reviewing the most recent research undertakings in PDS, the definition of PDS should be discussed, as well as the major components or elements of PDS and the service level. Definition of PDS The same problems encountered in attempting to define customer service occur in defining PDS. At one extreme, one could find gen- eral definitions that emphasize the time—and-place utilities provided by PDS, and at the other, definitions that either highlight the com- ponents or elements of PDS or establish performance standards to be attained by PDS. Another category of definitions can also be identi— fied, which stresses a supplier orientation or a customer orientation. The most general definition of PDS is the one that is closely related to the objective of PDS, i.e., creation of time-and-place utilities. Perreault and Russ (1976a) defined PDS as "the inter- related package of activities provided by a supplier which creates utility of time and place for a buyer and increases form utility" (p. 3). Heskett (1971) presented a list of definitions of PDS and ranked them in order of popularity. The definitions ranged from "the 16 elapsed time between the receipt of an order at the supplier's ware- house and the shipment of the order from the warehouse" to "the ease and flexibility with which a customer can place his order." The least-popular definitions are the customer-oriented ones, such as the latter above, in contrast to the supplier—oriented definitions, which seem to be more popular. The reason for this apparent paradox is the difficulties encountered by management in either controlling or measuring the customer-oriented performance levels. Customers are directly affected by the performance of customer service. Thus, a definition of customer service should account for the customer's needs, and PDS provides "the availability (at the right time and place) of a needed product“ (Perreault & Russ, 1974, p. 39). La Londe and Zinszer (1976, p. 271) summarized the three key ingredients associated with customer service definitions: (1) customer service is an activity that occurs at the customer-corporation inter- face; (2) customer service should involve only postsale activities; and (3) customer service is an evaluative measure, thus the perform- ance of customer service functions constitutes customer service. Since customer service is an evaluative measure, one can talk about service level, which constitutes performance standards for customer service functions. Thus, certain authors have presented the elements of PDS in terms of their performance measurements. PDS Elements It has been clear for some time that the major objective of physical distribution is to deliver finished products in the right 17 place, at the right time, in the right quantities and specifications, and in a usable condition. The elements or components of PDS should, theoretically, encompass all the functions and activities necessary to attain that overall physical distribution objective. In other words, they should comprise the functions or activities responsible for providing time-and-place utilities with assurance of form utility. An examination of the literature in the area of physical “"distribution showed some deviations from this theoretical framework when identifying the major components of PDS. Some authors have tended to include elements not related to the temporal-spatial utility-satisfaction objective. Appendix A contains a list of the major set of elements of PDS as perceived by various authors. Research in the Area of PDS After reviewing research in the area of PDS, it was concluded that a major category of components involves activities that influence the perspective of the buyer or have an effect on sales revenues: - PDS components and satisfaction with PDS components - PDS and the selection of suppliers - Causes of the importance of PDS in selecting suppliers Each of these topics is examined in detail in the following pages. PDS Components The majority of the research in this area is heavily concen- trated on order cycle time. The reasons for this emphasis seem to 18 be twofold: (l) the ease of measurement of order cycle time and (2) the availability of data on order cycle time. Ballou and DeHayes (1967) postulated that order cycle time variability was more important than average order cycle time in choos- ing transportation services. They conducted a simulation using speed of delivery, dependability (variations in delivery schedules), and the cost of service as variables for selecting service. The final outcome showed that dependability had an effect on inventory costs, whereas speed of delivery had little or no effect on the same costs, provided that the level of demand was kept constant. Aside from this stringent assumption, one cannot guarantee that the buyers will con- sider inventory costs when comparing average order cycle time and its variability. Hutchinson and Stolle (1968) conducted a survey of 500 customers of a firm to determine buyers' perceptions of the services provided. Among other results, it was found that delivery consistency was preferable to delivery time, and supplier inventory reliability was considered just as important as delivery time. The major short- coming of this survey rested in the sample used: It did not allow generalizations beyond the universe of the firm from which data were obtained. In a survey conducted by Ballou (1973), 2,000 members of the National Association of Purchasing Agents were selected to indicate change in supplier patronage if (1) average order cycle time were reduced, (2) order cycle time variability were reduced, or (3) if both were reduced. He concluded that "reliability may be less important 19 than we think" because buyers apparently cannot discriminate the effects of variability from the average order cycle time. The major problems of this research centered on methodological aspects (low response rates; nonrespondent bias) and also on the few aspects of PDS that were evaluated (Ballou also acknowledged this problem). In another simulation study, Speh and Wagenheim (1978) found that consistency in physical distribution performance was more impor- tant than "speed." Moreover, "inconsistent lead-times resulted in both higher system costs and lower customer service" (p. 106). The limitations of this study rested on the assumption that buyers would perceive the differences between consistency and "speed." Levy (1981) conducted an experiment with 30 drug wholesalers to define standards of customer service. From previous research, he determined the importance of components of customer service and found the following to be most significant: free order-placement policy (WATS line), terms of sale, consistent delivery, lead time, and fill rate. (It should be mentioned that terms of sale is ggt_a component of PDS according to the definition used in this study.) Levy con- tacted the subjects by telephone and asked them to rank nine dif- ferent pairings of service levels contained in five trade-off matrices of nine cells each. He examined the data using a conjoint analysis approach. The findings showed that the perceived dollar values of changes in lead time were greater than those of changes in consistent delivery. Generalizations from this study should be made with caution. The data-collection procedure was questionable, and 20 the rankings were made considering only three levels of customer ser- vice for each component. Uhr, Houck, and Rogers (1981) undertook a 43 factorial experi— ment to investigate the profit potential of certain customer service variables (order cycle time, variability of delivery time, and commu- nication time). In their model, profit was considered a linear func- tion of the customer service variables. The final outcome was that both order cycle time and variability of delivery time were signifi- cant, whereas communication time was not significant. However, any generalizations drawn from the study are questionable because some methodological problems could not be overcome, especially the non- response bias. The aggregate outcome of the above-mentioned studies is intriguing. From a theoretical standpoint, order cycle time varia- bility should be more important than average order cycle time, but the studies just reviewed showed evidence that contradicted the theo- retical interpretations. Concerning the importance of the PDS components, the La Londe and Zinszer study (1976) found the following order: product availa- bility (42.4), order cycle (20.7), distribution-system information (12.6), distribution-system flexibility (11.5), distribution-system malfunction (7.7), and postsale product support (4.5) (p. 117).1 These findings, however, are not representative of buyers' opinions because the number of respondents in this category (18) was too small 1The numbers in parentheses show the distribution out of 100 points given to that particular PDS element, indicating its importance. 21 to allow generalizations. Despite this shortcoming, it is worth mentioning that the rankings of importance varied widely by different industries. The researchers did not test for the significance of these differences. In a study conducted to evaluate the perceptions of the service supplied (order cycle time), Willett and Stephenson (1969) surveyed 480 drug and drug-sundry retailers. The researchers found that "buyers could discriminate among even small differences in physical distribution service times, and their ratings of satisfac- tion with service received were a linear function of service time" (p. 283). A major shortcoming of this study was that a recognition of a difference in service did not mean that the difference was significant. Concerning the perceived satisfaction of buyers with the service provided by their suppliers, Perreault and Russ (1976a) investigated 216 purchasing managers from rosters of the National Association of Purchasing Management. Respondents were asked to indicate on a seven-point rating scale their satisfaction with nine aspects of PDS. The findings showed a greater level of satisfaction with the aspects of P05 that dealt with communication (i.e., billing procedures, order methods, accuracy in filling orders). However, the elements that directly affected profits (delivery time varia- bility, and average delivery time) showed lower satisfaction levels. The problem with these findings was that satisfaction levels were only buyers' perceptions concerning the service provided. 22 Having examined the research studies on PDS components, a review of the importance of P05 in the patronage decision follows. PDS and the Selection of Suppliers This section contains a discussion of the research on the importance of PDS and the patronage decision. PDS and some of its components are contrasted with other factors that might affect the selection of suppliers from the point of view of the buyer. Klass (1961) interviewed 300 executives in 208 industrial companies to define what factors affect purchasing decisions. In order of importance, the most significant factors were: maintenance of product quality consistent with specifications on-time delivery performance an honest and sincere attitude on the part of salesmen price Note that a PDS component (delivery performance) was second only to product quality. In the Hutchinson and Stolle study (1968), delivery service was tied with product quality as suppliers' choice of the most important element. Lehmann and O'Shaughnessy (1974) undertook at investigation to determine "how the choice criteria used by purchasing agents to select suppliers vary with the type of problem likely to arise in adopting the particular product." The choice criteria were defined as the factors used to evaluate competitive offerings. The product categories were as follows: routine-order products (I), procedural- problem products (II), performance-problem products (III), and 23 political-problem products (IV). Table 2.1 summarizes some of their findings. It is worth mentioning that a PDS component (reliability of delivery) was an important factor in selecting suppliers. Table 2.1.--Rankings of selected factors by product type. Product Type Factors I II III IV Reputation 4 7 5 2 Flexibility 3 5 2 5 Technical service 12 l 3 7 Product reliability ll 11 4 3 Price 2 8 8 1 Reliability of delivery 1 4 l 4 Source: Lehmann and O'Shaughnessy (1974). Key: I = Routine-order products II = Procedural-problem products III = Performance-problem products IV = Political-problem products The major shortcomings of the study were of a methodological nature (Semon, 1975). The degree of influence of purchasing agents was not uniform among product classes or among different suppliers. Also, the ranking criteria were questionable; some of the differences between ranks may not have been significant. The sample size (45) was too small to allow further generalizations. In a study aimed at identifying the determinants of choice of supplier for capital goods in Great Britain, Cunningham and White (1973/1974) identified nine variables that influence the decision 24 process: price, delivery, reputation, past experience, technical specification, whether the machine was UK manufactured or imported, credit, trade-ins, and reciprocity. Only the first five variables had an effect in causing the buyer to discriminate between alternative offers. Price and delivery were extremely important for determining patronage decisions, provided that the technical specifications were met and that the reputation of the supplier for delivery reliability was not unfavorable. Cunningham and Roberts (1974) investigated the role of cus- tomer service in industrial marketing and concluded that "customer service can be a significant determinant of the buyer's attitude to his suppliers and on his eventual purchase decision" (p. 15). Per- sonal interviews were conducted with 25 buyers of steel castings and forgings among valve and pump manufacturers. The respondents were asked to rank 13 different customer service factors that could influ- ence supplier selection. They considered delivery reliability to be the most important factor. Some of the limitations of Cunningham and Roberts' findings should be mentioned. The sample size was very small, and thus no relevant statistical tests could have been made. Since only customer service factors were considered, other important factors for selecting suppliers were omitted. Banting (1976) replicated the preceding study in Canada. The sample size was larger (343 subjects), but the response rate was very low (73 usable responses), which imposed restrictions on the generalizability of the findings. In spite of the limitations, 25 delivery reliability was found to be the most important service factor in the purchase decision. Perreault and Russ (1976a) also explored the importance of certain supplier characteristics in the purchase decision. The various supplier attributes were measured using seven-point rating scales (ranging from "not important" to "very important"). The findings showed that product quality was perceived as the most impor- tant attribute, followed by distribution service and price. In the same study, Perreault and Russ explored some relation- ships that could exist between the importance of PDS and some situa- tional variables. For this purpose they formulated the following hypotheses (pp. 7-8): 1. "The greater the number of deliveries, the greater the importance of PDS. Problems with slow or inconsistent service could multiply with the number of deliveries." The findings confirmed the hypothesis. It is interesting to mention that the level of satisfac- tion with average delivery time and delivery-time variability were the lowest of all the PDS components. 2. "The greater the number of alternative suppliers avail- able, the lower the importance of PDS. Competition presumably increases efficiency, leading to a higher level of service and less need to pay attention to it in choosing suppliers." The findings were contrary to the hypothesis. 3. "The higher the proportion of cancelled orders, the greater the importance of PDS. Problems with PDS lead purchasers 26 to pay greater attention to it in choosing suppliers." The findings contradicted the original hypothesis. . 4. "The greater the satisfaction with PDS, the lower its importance in the evaluation. If the PDS the buyer is currently receiving is satisfactory, then he is less likely to consider it seriously in his decision process." The findings were consistent with the hypothesis, and the authors postulated that "the more important the service level, the more likely the purchaser is to seek out and patronize a supplier who meets his needs.“ 5. "The greater the average delivery time, the greater the importance of PDS." The findings did not show any apparent rela- tionship that could confirm the statement. A closer examination of these hypotheses and the findings presented by the authors led to the following conclusion: Other variables or factors affect the importance of PDS. Factors Affecting PDS Because customer service is measured by the performance of its components, there are certainly different levels of performance that should be adequate to meet specific service requirements. The paucity of research in this area makes it difficult to develop a concise frame of reference for analysis. La Londe and Zinszer (1976) concluded that "customer service is indeed situational to an industry and perhaps to the specific dis- tribution policies of a company" (p. 113). This statement consid- erably limits generalizations across industries in the area of PDS. 27 In line with this idea, Christopher et a1. (1977) emphasized that service is more "critical for some companies than for others" (p. 42). The major thrust is to determine if service is an important factor in influencing demand. Some factors are relevant to this under- taking: (1) product substitutability, which relates to brand loyalty and availability of close substitutes; (2) product criticality, which refers to the cost of a stockout; (3) complementary products; and (4) the cost of customer inquiries (p. 42). Heskett (1971) suggested observing the following factors in establishing service levels: (1) economics; (2) nature of the envi— ronment; (3) nature of the product, which includes substitutability and physical characteristics; and (4) pattern of demand. From the previous citations, one can infer that service levels will differ for every product and for each major market. Shycon and Sprague (1975) presented a host of factors to be considered when defining a certain service level. These included market share, frequency of purchase, customer's inventory policies, value of the product, profit margins, and degree of competition (p. 77). Framework of This Research As mentioned earlier, this research stemmed from the work done by Perreault and Russ (1976a). The hypotheses were spelled out in Chapter I and were very similar to the ones presented by Perreault and Russ (see page 6 of this dissertation), except for Hypothesis 2, in which backorders was used instead of cancelled orders, and Hypothe- sis 6 (company size and importance of PDS), for which Perreault and 28 Russ did not test. The content of the questionnaire resembled the one used by Perreault (1973) to generate the data base for the study he presented with Russ (1976a). The categories of variables were the following: Situational variables - number of deliveries"‘1 - percentage of backorders* - percentage of backorders cancelled* - average order cycle time* Supplier variables category of supplier number of suppliers utilized number of other suppliers number of alternative suppliers* Company variables - size by number of employees - size by sales volume - type of industry Satisfaction variables I .0 U U) components* billing procedures average delivery time delivery time variability rush service returns policy order status information accuracy in filling orders action on complaints - order methods - overall PDS* - Importance variables - PDS components2 1The variables marked with an * were the same as those pre- sented by Perreault and Russ (1976a). 2The components are identical to those presented for the satisfaction variables. 29 - Purchasing factors variables product quality* PDS* price* supplier management* distance to supplier* required order size* reciprocity* - Respondent variables - purchasing function - other functions or activities - general management finance production personnel marketing materials management - Product variable* - General variables attitudes toward possible stockouts changes in suppliers reasons for changing suppliers feedback on PDS* The situational, supplier, company, and satisfaction vari- ables were treated as independent variables in the hypotheses in which the purchasing-factor PDS was considered a dependent variable. For the other relational analyses, the situational, supplier, and company variables were the independent variables and the satisfaction, impor- tance, and purchasing-factor variables were the dependent variables. This chapter presented a review of the literature on customer service, with emphasis on PDS, and concluded with an outline of the basic framework of the research. In the next chapter, the basic pro- cedures used in this research are discussed. CHAPTER III RESEARCH METHODOLOGY Empirical data for this investigation were gathered by a survey of purchasing managers' opinions about the physical distribu- tion service (PDS) of industries located in two large industrial areas of the state of Rio Grande do Sul, Brazil. The data were collected in 1981, but respondents were asked to refer to their business activi- ties of 1980. In this chapter, a detailed description of the population surveyed, the data-collection procedures and instrument, the hypothe- ses, and the statistical tests is presented. Population Since purchasing managers play an influential role in the buying decision-making process, particularly on routine order products (Lehmann & O'Shaughnessy, 1974; Weigand, 1966), these managers were selected to be the source of information for this research. A preliminary contact was made with the newly established regional chapter (State of Rio Grande do Sul) of the Brazilian Asso- ciation of Purchasing Managers. Despite their interest in partici- pating in the study, two major problems arose: First, the membership roster had only 76 managers, and second, of these 76 members, only 12 could actually participate in the study. 30 31 Another alternative for reaching the purchasing managers was to obtain a list of all industries in the state of Rio Grande do Sul. This list was available in the Anuario das Indfistrias--l980, published by the Federacfio das Indfistrias do Rio Grande do Sul. Because this list contained a very large number of industries, several criteria were established so that the number of industries surveyed could be reduced. These criteria were: (1) size of the company (more than 50 employees); (2) geographical location of the company (either in the metropolitan area of Porto Alegre or in the city of Caxias do Sul--these two regions are the most industrialized areas of the state); and (3) type of industry (metallurgical, mechanical, electrical, transportation equipment, furniture, plastic, and shoes). The number of companies that met the above criteria was 418. A breakdown of the study population by industry and by region is presented in Table 3.1. Table 3.1.-—The study population, by industry and region. Industry Porto Alegre Caxias do Sul Total Metallurgy 144 28 172 Mechanical 30 12 42 Electrical 15 3 18 Transportation 7 5 12 Furniture 18 8 26 Plastic 23 5 28 Shoes 118 2 120 Total 355 63 418 32 Because it was feasible and desirable, all firms within the popula- tion were invited to participate in the study; thus, no sampling was necessary. In summary, the population of the study comprised purchasing managers of companies with more than 50 employees, located in either the metropolitan area of Porto Alegre or in the city of Caxias do Sul, both in the state of Rio Grande do Sul, and within a certain group of industries. Response Problems To avoid undesirable response rates that could lead to non- response errors, some precautionary actions were taken. These actions are briefly discussed below. Sponsorship of the Research There is evidence that if the person or organization sponsor- ing a particular research effort is made known to potential partici- pants, this tends to increase the response rate (Scott, 1961; Vocino, 1977). Moreover, Sponsorship by a university results in higher response rates than business-corporation sponsorship (Brunner & Carroll, 1969). Therefore, it was made clear to the respondents in the prenotification phone call and in the cover letter to the questionnaire that this investigation was sponsored by the Universi- dade Federal do Rio Grande do Sul. Prenotification Research has shown that prenotification, either by mail or by telephone, increases the response rate (Myers & Haug, 1969; 33 Waisenen, 1954). Two-thirds of the respondents in this investigation were contacted by phone before they received the questionnaire. Other measures that could improve the response rate were also followed. They were: including a pre-stamped return envelope (Ferris, 1951), assuring confidentiality, and setting a predetermined deadline for returning the questionnaire (Ferriss, 1951; Scott, 1961). Data-Collection Instrument The instrument used for collecting the data was a question- naire.1 The first page of the questionnaire gives general instruc- tions. The remainder of the instrument contains questions in the following areas: --selected product (Question 1) --supp1iers (Questions 2-4): category, number --company (Questions 5-7): size and type --respondent's activities in his/her company (Questions 8-9) --purchasing situations (Questions lO-l3): deliveries, back- orders, and order cycle time --evaluation of suppliers' PDS (Questions 14-23): satisfac- tion levels --importance of PDS components (Question 24) --factors influencing purchasing decisions (Question 25) --attitudes about specific purchasing situations (Questions 26-29) --general observations or comments (Question 30) 1The instrument was administered in Portuguese. The original questionnaire, including the cover letter and the English translation, is included in Appendix B. 34 Types of Products To assure more uniformity with reSpect to responses, the subjects were asked to answer questions in a product-specific situa- tion. That is, the respondents answered all the questions in reference to a specific product, chosen from a list provided in the general- instructions section of the questionnaire. The products included in that list met the following criteria: (1) routine order products, (2) repeated purchases, (3) various suppliers, and (4) high usage among different industries. The investigator chose the following products for inclusion in the list: fasteners, bearings, lubricants, abrasives, electrodes, and acids. Test of the Instrument The preliminary version of the questionnaire was tested with a group of purchasing managers of eight different companies. The major problems with the questionnaire centered on its format and length, the difficulties of coding (in the first version, the respondents were supposed to do the coding themselves), and the understanding of the term "logistics," with which the managers were not familiar. Appropriate adjustments were made to accommodate the suggestions provided by the respondents. The format and length were reduced and the cover letter included in the body of the question- naire. Some questions were condensed and others eliminated. The coding was omitted from the questionnaire, and the term "logistical services" was substituted for "distribution services." 35 Data-Collection Procedures The data collection followed five distinct phases, which are summarized below: 1. Selection of industries and then companies from the Anuario das Indfistrias--l980. 2. Telephone contact with each company, to confirm the address and to obtain the name of the purchasing manager or the per- son responsible for the purchasing function. 3. Telephone conversations with the purchasing managers to explain the research theme and to obtain from them a commitment to complete the questionnaire to be sent the following week. It was emphasized that the research was sponsored by the Universidade Federal do Rio Grande do Sul and that all data would be kept confi- dential. Even though the researcher had intended to contact all the respondents by telephone, two problems jeopardized this objective: first, some managers could not be reached; second, cost restrictions prevented telephone contacts with the respondents from the Caxias do Sul area. Thus, of the 355 subjects in Porto Alegre, 276 were contacted (78%). The response rates were as follows: Porto Alergre (telephone contact with purchasing manager) -- 46.0% Porto Alegre (telephone contact with company) -- 38.0% Caxias do Sul (no telephone contact) -- 36.5% 36 4. Mailing of the questionnaire with a pre-stamped return envelope. 5. Personal contact with 22 nonrespondents to obtain their response to the questionnaire. Despite efforts to reduce the non- response rate, 238 individuals failed to return the questionnaire (see Table 3.2 for details). To minimize nonresponse error, a standard procedure was used. A random sample was drawn from the nonrespondent population (10%). The researcher contacted each of these nonrespond- ents (i.e., the purchasing manager) by telephone, emphasizing the theme of the study and the importance of his/her participation. Also, a personal appointment was scheduled with the nonrespondent for the purpose of completing the questionnaire. During this appointment, the investigator was careful to avoid influencing the responses the subject might give. A total of 22 questionnaires were obtained using this procedure (2 of the sample of 24 nonrespondents failed to com- plete the questionnaire). 6. Test for nonresponse error. To allow the research find- ings to be generalized to the entire population, tests were conducted to compare differences between the subjects who responded to phase- four stimuli and those who responded to phase five stimuli. For statistical purposes, the respondents were divided into two groups: (1) mail group and (2) personal-contact group. Also, the responses were classified in three categories: (1) dichotomies (variables 8-14 and 46); (2) nominal or ordinal responses (variables 1, 2, 5-7, 19-45, and 47-49); and (3) interval or ratio-scaled responses (variables 3, 4, and 15-18). (See Appendix C for the list of variables.) Table 3.2.--Questionnaire responses, by industry and location. 37 Questionnaires Industry Not Personal Total Returned Usable Returned Contact Usable Porto Alegre Metallurgy 66 64 78 7 71 Mechanical 13 13 17 2 15 Electrical 6 6 9 - 6 Transportation 2 2 5 - 2 Furniture 11 9 7 1 10 Plastic 12 9 11 l 10 Shoe 47 43 71 7 50 Total 157 147 198 18 165 Caxias do Sul Metallurgy 10 9 18 1 10 Mechanical 5 5 7 l 6 Electrical 1 1 2 - 1 Transportation 2 2 3 - 2 Furniture 2 2 6 l 3 Plastic 2 2 3 1 3 Shoe l 1 l - 1 Total 23 22 40 4 26 Grand total 180 169 238 22 191 38 The dichotomies were tested using a standard chi-square pro- cedure. The differences between the two groups were not statis- tically significant at a = .05. But at a = .10, the two groups were different with respect to responses on variable 10 (financial activi- ties). A t-test was used to examine the differences between the two groups for the interval or ratio-scaled responses. None of the dif- ferences was statistically significant at a = .05, nor at a = .10. For the nominal or ordinal responses, a test of differences of proportions, recommended by Fisher (Guilford, 1965), was used. Since in both groups the sample size was not small (<10), the sampling distribution of the difference in the proportion approached normality. Therefore, the test of significance was made through use of a 7 ratio. Three differences were significant: importance of delivery time variability (rank 1), importance of accuracy in filling orders (ranks 1, 4, and 6), and importance of PDS (ranks 3 and 5). Therefore, there appeared to be no statistical differences (with the exceptions mentioned) between the mail group and the personal-contact group of respondents. That is, it was not necessary to account for nonresponse error. Research Objectives and Methodology The objectives of this research undertaking were discussed in Chapter I. In this chapter the research objectives are examined from a methodological perspective. That is, every objective is related to the questionnaire design. The statistical hypotheses and the tests used to verify them are also presented. 39 Objective 1: Satisfaction To investigate the perceived satisfaction of buyers with overall PDS and with each of its components. Questions 14-22 in the questionnaire deal with respondents' satisfaction with PDS components, and Question 23 is a rating of their overall satisfaction with PDS. As can be seen on page 4 of the ques- tionnaire, the respondents were asked to rate their satisfaction with PDS and its components on a five-point scale, ranging from totally dissatisfied to totally satisfied. For purposes of measurement, the ratings were scaled from 1 (totally dissatisfied) to 5 (totally satisfied). Therefore, the data were measured on an ordinal scale. Scales of this type might have some limitations concerning the statistical tests that can be performed. The basic limitation stems from the failure to meet the equal-interval postulate of measurement. However, as Kerlinger (1973) pointed out, "yet most [behavioral] scales are basically ordinal, [and] we can with considerable assurance often assume equality of interval" (p. 440). Satisfaction ratings were averaged to permit ranking of the perceived satisfaction with the various PDS components. To test for the statistical differences of the mean ratings between the vari- ables, the Wilcoxon matched-pairs signed test was used. This test allows the researcher "to tell which member of a pair is greater" (Siegel, 1956, p. 75). Thus, 36 signed tests were performed. One word of caution should be given concerning the level of significance in this type of test. The a values are used for every pair comparison, 40 but they are not additive. That is, there is no significance level for the overall ranking of the mean ratings. Satisfaction with the overall PDS was measured in two differ- ent ways. The first was identical to the one used for the PDS com- ponents (ratings ranging from totally dissatisfied to totally satis- fied). The second was a weighted average of the components ratings. The weights were determined by the importance ranking of every PDS component given in Question 24. Parametric correlations were run between every rating of satisfaction and the corresponding ranking of importance to test for independence. In only two cases was the correlation coefficient significant: for "rush services" and for "order status information." However, both coefficients were very small (.25 and .12, respectively) but negative, indicating the fol- lowing pattern: the higher the satisfaction, the higher the impor- tance of the component. Objective 2: Impgrtance To rank, in order of importance, different components of PDS as perceived by purchasers. Question 24 was included to find out about the importance ranking of the PDS components. To allow for better discrimination, the respondents were asked to rank the different components of PDS in order of importance, rather than using a Likert-type scale, which would have rated each component in terms of its importance (i.e., from not very important to very important). 41 For measurement purposes, the rankings were scaled from 1 to 5, in descending order of importance, producing an ordinal-type scale. Some coding adjustments were made to deal with ties and with nonresponses. To compare every pair of rankings, a Wilcoxon signed test was again used. The same comments mentioned in the previous section applied to the tests conducted here. Objective 3: Importance To compare the importance of PDS with other factors influencing patronage decisions. To evaluate the importance of different factors affecting the purchasing decision, Question 25 was included in the question- naire. Respondents were asked to rank, in order of importance, the various factors that could influence their purchasing decisions. The responses were measured on an ordinal scale, with l for the most important factor and 7 for the least important. Again, this measurement was preferred to the Likert-type scale because it allows better discrimination of responses. To test for ranking differences between the various factors, the same Wilcoxon paired-comparison test was performed. The rankings were only an indication of the perceived importance of any one of the factors. In no way were they manifestations of or surrogates for actual purchasing behavior. 42 Objective 4: Relationships To explore relationships between situational variables of the buying process, supplier vari- ables, company variables, and satisfaction vari- ables with the perceived importance of PDS in selecting suppliers. This objective, as presented in Chapter I, was operationalized in the form of hypotheses. In this section, each hypothesis is examined and discussed from a methodological perspective. Moreover, additional relationships explored in the research are also discussed here. Situational variables.--The situational variables1 were defined as follows: --number of deliveries (Question 10, variable 15) --percentage of backorders (Question 11, variable 16) --percentage of backorders cancelled (Question 12, variable 17) --average order cycle time (Question 13, variable 18) Hypothesis 1: The greater the number of deliveries, the greater the importance of PDS in selecting suppliers. To test this hypothesis, the Pearson correlation coefficient between number of deliveries (Question 10, variable 15) and the importance of PDS (Question 25, variable 39) was employed. The sta-‘ tistical and the null hypotheses that resulted from the substantive hypothesis were, respectively: Ho: r 3 0 where r = correlation coefficient between number of deliveries H]: r < O and importance of PDS 1Refer to Appendix C for a complete list of variables 43 Thus, if the Pearson correlation coefficient was significantly different from zero and negative, the null hypothesis should be rejected and the alternative (or statistical) hypothesis accepted. Additional tests concerning the hypothesized relationship were performed using univariate analysis of variance. The variable number of deliveries was divided into three groups according to the frequency of deliveries per month. The importance of PDS was the dependent variable. This procedure enabled the researcher to examine whether the three groups were different in terms of their responses to the ranking of PDS importance. If the F-value was not statistically significant, the null hypothesis of no difference could not be rejected. Hypothesis 2: The higher the proportion of backorders, the greater the importance of PDS in selecting suppliers. Again the Pearson correlation coefficient was used to test this hypothesis about the relationship between proportion of back- orders (Question 11, variable 16) and the importance of PDS.1 The testable hypotheses were: Ho: r 2 0 where r = correlation coefficient between proportion of back- H]: r < 0 orders and importance of PDS The acceptance conditions were identical to the ones presented above for Hypothesis 1. 1The proportion of backorders cancelled could also have been tested with respect to the importance of PDS, but response problems on Question 17 prevented the tests from being conducted. 44 Because a considerable number of respondents reported no backorders, this variable was divided into two groups, one with no cancelled backorders and the other with one or more cancelled backorders. An analysis of variance, in this case the standard t-test since only two groups existed, was performed to examine dif- ferences with respect to responses to importance of PDS. Hypothesis 3: The greater the average order cycle time, the greater the importance of PDS in select- ing suppliers. This hypothesis involved two variables: average order cycle time (Question 13, variable 18) and importance of PDS. The statis— tical tests were similar to the ones presented for both Hypotheses l and 2. That is, H0: r 2 0 where r = correlation coefficient between order cycle time H]: r < 0 and importance of PDS Supplier variables.--The supplier variables were defined as follows: --category of supplier (Question 2, variable 2) -—number of suppliers utilized (Question 3, variable 3) --number of other suppliers (Question 4, variable 4) --number of alternative suppliers (variable 50 = V3 + V4) Hypothesis 4: The greater the number of alternative sup- pliers available, the lower the importance of PDS in selecting suppliers. The variables that were dealt with in examining this hypothe- sis were number of alternative suppliers and importance of PDS. The testable hypotheses were: 45 H0: r g 0 where r = correlation coefficient between number of alterna- H]: r > O tive suppliers and impor- tance of PDS The null hypothesis was rejected if the Pearson correlation coeffi- cient was significantly different from zero and positive. Company variables.--This group of variables was defined as: --company size, in terms of number of employees (Question 5, variable 5) --company size, in terms of sales volume (Question 6, variable 6) --company's industry (Question 7, variable 7) Hypothesis 5: The larger the company, either in terms of number of employees or in sales volume, the greater the importance of PDS as a factor in selecting suppliers. Because three variables were involved in this hypothesis (size in terms of employees and sales volume and importance of PDS), one could have two sets of testable hypotheses: Ho: r 3 0 where r = correlation coefficient between size (number of H]: r < 0 employees) and importance of PDS H6: r' 3 0 where r'= correlation coefficient between size (sales volume) H3: r' < 0 and importance of PDS Satisfaction variables.-—These variables were defined as follows: --components of PDS (Questions 14-22, variables 19-27) --overall PDS (Question 23, variable 28 or weighted average of PDS components, variable 51) 46 Hypothesis 6: The greater the satisfaction with PDS, the lower its importance as a purchasing factor. This hypothesis also was tested using the Pearson correlation coefficient for the following: Ho: r 2 0 where r = correlation coefficient between satisfaction with H]: r < 0 overall PDS and importance of PDS Objective 5: Comparative Analysis To present a comparative analysis of the findings of this study with those reported by Perreault and Russ (1976a). This analysis compares the findings in the Brazilian and Ameri- can environments on the following aspects: (a) satisfaction with overall PDS and each of its components, (b) the importance of PDS as a factor influencing patronage decisions, and (c) relation- ships involving the perceived importance of PDS in selecting suppliers. The procedures used in making the comparisons were straight- forward, but before each one is discussed, a word of caution is sug- gested concerning the measurements. The satisfaction ratings were obtained on different scales: for this research, on a five-point scale ranging from totally satisfied to totally dissatisfied; and for the Perreault and Russ study, on a seven-point scale ranging from satisfactory to unsatisfactory. Ratings of the importance of factors in selecting suppliers were also gathered in different ways: for this research, on a ranking of importance from the most important to the least important; and in the Perreault and Russ study, from not impor- tant to very important. Thus, comparisons cannot be made in an abso- lute manner, but only on a relative basis. 47 Satisfaction.--The comparisons were made on three different levels. The first examined the order of satisfaction ratings of the PDS components. The second compared the correlations of each of the components with overall PDS. And the third discussed the findings with respect to feedback of service (Question 29 of the questionnaire: "Do your suppliers of this product check with you to see if the ser- vices they are providing are adequate in meeting your needs?"). Importance.--In this comparative aspect, the findings of the two studies were contrasted on the basis of the rank order of impor- tance of factors in selecting suppliers. Relationships.--The comparisons were made using the hypotheses outlined in the last section, which dealt with situational variables of the buying process (number of deliveries and order cycle time), a supplier variable (number of alternative suppliers), and satisfaction with overall PDS. To allow for more meaningful analyses, all the means of each of the above variables were broken down by every rank of importance of PDS in selecting suppliers. Other Tests To provide a sharper picture of the findings, other relation- ships, associated with the objectives, involving supplier variables, situational variables, and company-size variables were examined. Supplier variables.--The Pearson correlation coefficient was used to determine the existence of relationships between the number of alternative suppliers and satisfaction with overall PDS (measured either by Question 23 or by the weighted average of PDS components). The hypotheses were of the following type: 48 H0: r = 0 where r = correlation coefficient between number of alternative H]: r f 0 suppliers and satisfaction with overall PDS Situational variables.--An examination of the relationships between number of deliveries, percentage of backorders, and order cycle time and the level of satisfaction with overall PDS (measured either by Question 23 or by the weighted average of the PDS components) was performed. The hypotheses to be tested were of the following type: H0: r = 0 where r = correlation coefficients between situational variables H]: r f O and satisfaction with overall PDS A t-test was conducted to explore further the relationships between situational variables and satisfaction with P05 (in this case, measured by the weighted average of PDS components). Specifically, the variable percentage of backorders was divided into two groups: one had no backorders and the other had one or more. The tested hypothesis was straightforward: no difference in responses to PDS satisfaction between the two groups. Company variables.--Several additional relationships were explored using the two company-size variables. First, they were related to overall satisfaction with PDS (measured either by Question 23 or by the weighted average of PDS components). The hypotheses examined were of the following pattern: Ho: r = 0 where r = correlation coefficients between company size and H]: r f 0 overall satisfaction with PDS 49 Second, the importance of every PDS component was correlated with both company-size variables. These relationships were tested with the following types of hypotheses: Ho: r = 0 where r = correlation coefficients between company size (employ- H]: r f 0 ees and sales) and the impor- tance of every PDS component Third, univariate analyses of variance were used to test the difference between the various company-size groups, given by number of employees and sales volume, with respect to their rankings of impor- tance of PDS components and of factors for selecting suppliers. Given the two independent variables (company size), there were 18 ANOVAs for the PDS components (2 x 9) and 14 ANOVAs for the factors for selecting suppliers (2 x 7). The 32 different hypotheses had the following format: HO: O] = p2 = ... = pk = p where p's = means for the dif- erent groups H]: u] f uz f -.- f pk r u Fourth, multivariate analyses of variance were used to examine the differences among diverse company sizes and the aggregate responses for the following sets of variables: satisfaction with PDS components, importance of PDS components, and factors in selecting suppliers. Therefore, six different MANOVAs were performed, as shown below: MANOVA GROUPS DEPENDENT VARIABLES A size--employees satisfaction B size--employees importance C size-~employees factors in selecting suppliers D size—-sales satisfaction E size--sa1es importance F size--sales factors in selecting suppliers 50 The multivariate F-statistic was used to test the significance of the differences among the various groups on all the MANOVAS. If the differences among groups were significant, one could compute the univariate F-statistics to determine which variable or variables were relevant in explaining the differences. Fifth, standard ANOVAs were used to explore further the relationship between company size and overall satisfaction with PDS. Groups were defined for both measurements of company size. Four ANOVAs were instrumental for testing four hypotheses of this sort: H : p] = p2 = ... = pk = p where p's = means for the dif- ferent groups H]: u] f Hz 7 --- f uk r u The procedures described in this chapter created the data- file source from which the relational analyses and findings of the study were developed. The next chapter contains a discussion of these findings. CHAPTER IV RESEARCH FINDINGS This chapter contains major findings and analyses that should lead to the fulfillment of the purpose and objectives of this investi- gation. In the first section, the general descriptive findings are presented to provide a frame of reference for the relational findings. Following are sections that center on satisfaction with PDS, importance of PDS components, factors affecting the patronage decision, and PDS and the patronage decision. These sections provide the background for the comparative analysis that follows. A summary of the findings that most directly relate to the research objectives and hypotheses is given in the last section. General Descriptive Findings In this section, data on the respondents, the suppliers, the chosen product, and the purchasing situation are presented to provide a general overview of the descriptive findings of this research. These data consist mainly of frequency distributions, percentages, and cumu- lative percentages. Respondents The findings about the respondents may be grouped in two cate- gories. The first deals with the respondents' activities within the firm and the second with some characteristics of their companies. 51 52 Respondents' activities.--Even though the subjects of this research were purchasing managers within a certain group of indus- tries, not all respondents dealt solely with purchasing in their companies. As can be seen in Table 4.1, almost half (46.6%) of the respondents had other activities besides purchasing. Table 4.1.--Activities of the respondents. Activity N % Only purchasing 102 53.4 Purchasing and other activities 89 46.6 Total 191 100.0 Table 4.2 shows that the "other" activities encompassed almost all of the managerial functions. Production and materials management rep- resented 53.5% of the other activities. Mentions of personnel, finance, and marketing could have been a result of the small size of some of the companies. Table 4.2.--Distribution of "other" activitiesa of the respondents. Activity N % General management 12 10.5 Finance 20 17.5 Production 27 23.7 Personnel 4 3.5 Marketing 17 14.9 Materials management 34; 29.8 Total 114D 100.0 aCorresponds to the "purchasing and other" category from Table 4.1. bTotal greater than 89 because of multiple responses. 53 Characteristics of the respondents' companies.--This investi- gator did not intend to evaluate or to present a host of different characteristics of the companies researched. However, some character- istics relating to size and also to the type of industry were identi- fied. This was done, first, to give a clearer picture of the companies and, second, to permit further analyses of relationships of company size to satisfaction with and importance of PDS components and to pur- chasing factors. In Table 4.3, company size is measured by number of employees. Almost 80% of the companies had fewer than 500 employees. Only 1% of the companies had more than 5,000 employees. Table 4.3.--Distribution of respondents by company size (# of employees). Number of Employees N % Cumulative % Less than 50 8 4.2 4.2 50 to 99 51 26.7 30.9 100 to 499 92 48.2 79.1 500 to 999 16 8.4 87.5 1,000 to 4,999 22 11.5 99.0 5,000 or more 2 1.0 100.0 Total 191 lOOJOF -- When company size was measured in terms of sales volume (see Table 4.4), about three-quarters (73.4%) of the companies showed sales volume figures below 500 million cruzeiros. The industry categories of the respondents' companies are shown in Table 4.5. The majority of the companies were either metal- lurgical (23.8%), mechanical (23.8%), or shoe industries (27.6%). 54 Although all of the companies were officially classified within the industry groups selected for this study, 17 respondents (9.2%) clas- sified their firms as belonging to other industries. Table 4.4.—~Distribution of respondents by company size (sales volume). Sales Volumea N % Cumulative % Less than 49 27 14.7 14.7 50 to 99 31 16.8 31.5 100 to 499 77 41.8 73.4 500 to 999 20 10.9 84.2 1,000 or more 29 15.8 100.0 No response 7 -- -- Total 191 100.0 -- aValues in millions of cruzeiros in 1980. Table 4.5.--Distribution of respondents by type of industry. Type of Industry N % Metallurgy 44 23.8 Mechanical 44 23.8 Electrical 9 4.9 Transportation equipment 4 2.2 Furniture 7 3.8 Plastic 9 4.9 Shoes 51 27.6 Other 17 9.2 No response 6 -- ...: O O C Total 191 55 Suppliers Four major variables were analyzed with respect to suppliers. They included: number of suppliers utilized, number of other suppliers, number of alternative suppliers (that is, suppliers utilized plus other suppliers), and supplier category. Each variable is discussed below. Number of suppliers utilized.--A distribution of the number of suppliers utilized by the purchasers' companies is shown in Table 4.6. These were the suppliers with whom the respondents' companies actually did business in 1980 with respect to the chosen product. From the table, it can be seen that close to 90% of the companies used fewer than five suppliers. Only 2.7% had more than 10 active suppliers for the chosen product. Table 4.6.--Distribution of number of suppliers utilized. Number of Suppliers N % Cumulative % l 31 16.9 16.9 2 46 25.1 42.0 3 53 29.0 71.0 4 23 12.6 83.6 5 11 6.0 89.6 6 to 10 14 7.7 97.3 11 to 34 5 2.7 100.0 Missing 8 -- -- Total 191 100.0 -- Number of other suppliers.--To determine the total number of suppliers available to their companies, the respondents were asked to mention the number of other suppliers with whom they could have done business. The summary of the responses is presented in Table 4.7. 56 Table 4.7.--Distribution of number of other suppliers available. Number of Suppliers N % Cumulative % O 4 2.4 2.4 1 11 6.7 9.1 2 16 9.8 18.9 3 23 14.0 32.9 4 25 15.3 48.2 5 20 12.2 60.4 6 to 10 43 26.2 86.6 11 to 50 22 13.4 100.0 Missing 27 -- -- Total 191 100.0 -- Number of alternative suppliers.--The number of suppliers uti- lized and the number of other suppliers were added to give the total number of alternative suppliers. The distribution of these suppliers is shown in Table 4.8. Note that almost three-quarters (73.4%) of the respondents' companies could have used from 1 to 10 suppliers. Table 4.8.--Distribution of number of alternative suppliers. Number of Suppliers N % Cumulative % l 4 2.2 2.2 2 8 4.3 6.5 3 18 9.8 16.3 4 19 10.3 26.6 5 15 8.2 34.8 6 15 8.2 43.0 7 19 10.3 53.3 8 17 9.2 62.5 9 15 8.2 70.7 10 5 2.7 73.4 11 to 20 39 21 2 94.6 21 to 64 10 5.4 100.0 Missing 7 -- -- ...o o O O I I Total 191 57 Supplier category.--The most important suppliers for the respondents' companies were classified in three categories: manu- facturer, wholesaler, or distributor. The majority of the suppliers (65.3%) were classified as manufacturers. Table 4.9 presents the breakdown. Table 4.9.--Supplier categories. Category N % Manufacturer 124 65.3 Wholesaler 37 19.5 Distributor 29 15.2 No response 1 -- Total 191 100.0 Purchasing Situation The purchasing situation involved four different variables. They were: number of deliveries, backorders, backorders cancelled,1 and average order cycle time. All the responses concerned transac- tions conducted in 1980 and were product and supplier specific. The descriptive findings for each variable are presented below. Number of deliveries.—-To provide a better picture of the number of deliveries, the data were grouped in four categories, as shown in Table 4.10. More than two-thirds (68.3%) of the respondents had, on the average, three or less deliveries per month in 1980, and 1Because of response problems for question 17, which dealt with proportion of backorders cancelled, no findings are reported concerning this variable. 58 close to 30% had 12 or less deliveries in 1980, representing one or less deliveries, on the average, per month. Table 4.10.--Distribution of number of deliveries in 1980. Number of Deliveries % m i Per Year Per Month3 N / Cu ulat ve % 3 to 12 1 or less 43 29.7 29.7 13 to 24 1 to 2 38 26.2 55.9 25 to 36 2 to 3 18 12.4 68.3 37 to 120 3 to 10 46 31.7 100.0 Missing 46 -- -- Total 191 100.0 -- aPer month equals per year divided by 12. Backorders.--The data in Table 4.11 show a distribution of the percentages of backorders in 1980 for the chosen product. The majority of respondents (54.3%) did not have backorders; 78% of them had 5% or fewer of their orders backordered. Table 4.11.--Distribution of percentage of backorders. Percentage of Backorders N % Cumulative % 0 96 54.3 54.3 1 8 4.5 58.8 2 3 1.7 60.5 3 10 5.7 66.2 4 2 1.1 67.3 5 19 10.7 78.0 From 6 to 20 28 15.8 93.8 From 21 to 70 11 6.2 100.0 Missing 14 -- —I O 0 : O Total 191 59 Average order pycle time.--The data on the average order cycle time were arranged in various intervals to provide a more concise pic- ture of the distribution of the responses. As can be seen in Table 4.12, the majority of the respondents had average order cycle times of 30 days or less, and only 5.4% had average order cycle times of more than 60 days. Table 4.12.--Distribution of the average order cycle time. Average Order Cycle Time (1n days) N % Cumulative % 1 to 5 44 26.2 26.2 6 to 10 31 18.4 44.6 11 to 20 32 19.0 63.6 21 to 30 25 14.9 78.5 31 to 45 8 4.8 83.3 46 to 60 19 11.3 94.6 61 to 90 9 5.4 100.0 Missing 23 -- -- Total 191 100.0 -- The Chosen Product In the methodology chapter, it was emphasized that the ques- tionnaire responses were product specific. That is, from a list of six products, the respondents were asked to choose one product and to answer the questions considering transactions in relation to that product. In Table 4.13 the distribution of the chosen products is presented. Fewer than 10% of the respondents did not select a product from the list provided in the questionnaire. Since the questions were product specific, one might wonder about differences in responses of the subjects who chose different products. But in conducting a series of statistical tests (ANOVAs 60 for the ordinal or interval-scaled responses and crosstabs for the dichotomies), no major differences were found. Table 4.13.--Distribution of the chosen products. Product N % Fasteners 36 18.9 Bearings 44 23.2 Lubricants 21 11.1 Abrasives 40 21.1 Electrodes 18 9.5 Acids 14 7.4 Others 13 6.8 Several 4 2.1 No response 1 -- Total 191 100.0 Other General Descriptive Findings Three more general descriptive groups of findings were derived from the questionnaire. The first involved the supplier obtaining feedback from the purchaser about the PDS provided. The second out- lined respondents' possible actions in a stockout situation. The third concerned the reasons for changing suppliers. Feedback about services. Figure 4.1 is a representation of the responses given by the purchasers about the feedback that suppliers seek concerning the services provided. The majority of the respondents (l42--81%) said that their suppliers contacted them to check on the adequacy of the services. Of these 142 subjects, 68% (103 cases) recognized the need for improvement; in fact, in 89% of the cases the services did improve. 61 Yes N=92 89% Did services N:?35'improve? 68% No Yes__Improvements III] N=152 necessary? 81% No N=49 . 32% Supplier Seeking Feedback About Services N=l88 Yes N=24 No--Wou1d like 67% N=36 them to 19% do it? No N=12 33% Figure 4.1.--Suppliers seeking feedback about physical distribution services. Actions in a stockout situation.--The respondents were asked to mention their possible action in a stockout situation by their most important supplier of the chosen product. A large number of the sub- jects would change suppliers (142--83%), but 54% (77) of them just for that specific order; only 4% (6) would change suppliers perma- nently. Figure 4.2 is a diagram of the various responses given by the subjects. 62 Permanently N=6 (4%) For this order N=77 (54%) Change sup lier N=142 (83%) Only if order is urgent N=ll (8%) Only if no .other arrangements can be made N=48 (34%) STOCKOUT Can wait N=l72 N=lO (6%) Supplier takes appropriate action N=ll (6%) Other N=9 (5%) Figure 4.2.-—Actions concerning a stockout possibility. Changing;suppliers.--Table 4.14 is a summary of the responses to a question concerning changes in suppliers for the chosen product in the last two years. Nearly 44% of the respondents had changed suppliers; 56.5% had not. Responses to a question concerning the reasons for changing suppliers are summarized in Table 4.15. The major reason for change was related to price (35.2%), followed by problems with PDS (23.2%) and then problems with product quality (24.0%). 63 Table 4.14.--Changes of suppliers in the last two years. N % Yes 81 43.5 No 108 56.5 No response 2 -- Total 191 100.0 Table 4.15.--Distribution of major reasons for changing suppliers. Reason N % Price 44 35.2 Product quality 29 23.2 PDS 30 24.0 Other 22 17.6 Total 125a 100.0 aTotal greater than 81 because of multiple responses. Satisfaction With PDS and Its Components In this section, findings concerning the satisfaction with PDS components are analyzed in two ways. The first deals with purely descriptive aspects of the findings. The second explores relation- ships among the PDS components to provide some insights in attempting to explain the descriptive findings. Descriptive Findings_ The respondents rated their satisfaction with each PDS com- ponent and with the overall PDS on a five—point scale ranging from totally dissatisfied (1) to totally satisfied (5). In Table 4.16 64 the means of these ratings are presented, as well as the correspond- ing standard deviations. Table 4.16.--Satisfaction ratings for PDS components. Mean Satisfaction Standard PDS Components Rating Deviation Billing procedures 3.842 .877 Average delivery time 4.021 .831 Delivery time variability 3.807 .858 Rush services 3.856 1.092 Returns policy 3.896 .911 Order status information 3.978 .878 Accuracy in filling orders 4.058 .849 Actions on complaints 4.124 .758 Order methods 4.134 .646 Examining the deviations in Table 4.16 and relating them to the means, one might question the possibility of ranking the PDS components according to their respective satisfaction rating. Never- theless, a Wilcoxon matched-pairs signed test was performed for every difference between two means of satisfaction ratings. In Table 4.17, each of the tests is presented along with the z-values, which are approximately normally distributed with zero mean and unit variance (Siegel, 1956, p. 79), and the p-values associated with each z-value. Examining the mean ratings given in Table 4.16 and the various pair- comparisons presented in Table 4.17, the PDS components can be ranked according to their satisfaction ratings, as follows (at a p value < .10 for every pair comparison): Table 4. 65 l7.--Summary of the Wilcoxon test for mean ratings of satis- faction for PDS components. Pair of Variables z—valuea p-value Billing procedures/average delivery time 1.970 .024 Billing procedures/delivery time variability -.367 .357 Billing procedures/rush services .065 .474 Billing procedures/returns policy .916 .180 Billing procedures/order status information 1.762 .039 Billing procedures/accuracy in filling orders 2.590 .005 Billing procedures/actions on complaints 3.852 .000 Billing procedures/order methods 4.113 .000 Average delivery time/delivery time variability -3.909 .000 Average delivery time/rush services -2.453 .007 Average delivery time/returns policy -l.682 .046 Average delivery time/order status information -.463 .322 Average delivery time/accuracy in filling orders .528 .299 Average delivery time/actions on complaints 1.817 .035 Average delivery time/order methods 1.816 .035 Delivery time variability/rush services .398 .345 Delivery time variability/returns policy 1.013 .155 Delivery time variability/orderstatusinfbrmation 2.407 .008 Delivery time variability/accuracy in filling orders 3.505 .000 Delivery time variability/actions on complaints 4.755 .000 Delivery time variability/order methods 4.670 .000 Rush services/returns policy .526 .299 Rush services/order status information 1.408 .080 Rush services/accuracy in filling orders 2.453 .007 Rush services/actions on complaints 3.487 .000 Rush services/order methods 3.257 .001 Returns policy/order status information .869 .192 Returns policy/accuracy in filling orders 2.084 .019 Returns policy/actions on complaints 3.518 .000 Returns policy/order methods 3.333 .000 Order status information/accuracy in filling 1 012 156 orders ' ' Order status information/actions on complaints 2.431 .008 Order status information/order methods 2.517 .006 Accuracy in filling orders/actions on complaints 1.483 .069 Accuracy in filling orders/order methods 1.386 .083 Actions on complaints/order methods .103 .459 az=T-]JT 0'r where T = the smaller sum of like-signed ranked. 66 --Group 1 (higher satisfaction rating): order methods and action on complaints. At an a value of .10, the means of these com- ponents were significantly larger than the mean of any other component. However, the difference between the means of order methods and actions on complaints was not significant (p = .459). --Group 2: accuracy in filling orders and average delivery time. At an a level of .10, their means were smaller than the means of order methods and actions on complaints and larger than the means of returns policy, rush services, billing procedures, and delivery time variability. However, they were not different from the mean of order status information, nor were they different among themselves. Since the means of returns policy and order status information were not significantly different and because the mean of returns policy was significantly different from the means of both accuracy in filling orders and average delivery time, the component order status informa- tion was omitted from group 2 and constituted group 3. --Group 3: order status information. At an a level of .10, the mean of this component was smaller than the means of accuracy in filling orders and average delivery time and was larger than the means of rush services, billing procedures, and delivery time variability. Because of these differences and the fact that the mean of returns policy was not different from the means of rush services, billing procedures, and delivery time variability, the component returns policy constituted group 4 and the components rush services, billing proce- dures, and delivery time variability formed group 5. --Group 4: returns policy. 67 --Group 5 (lower satisfaction rating): rush services, bill- ing procedures, and delivery time variability. The mean rating for satisfaction with the overall PDS was 3.958, with a standard deviation of .778. The distribution of the ratings for the overall PDS is presented in Table 4.18. As can be seen in this table, the majority of respondents (85.7%) said they were either satisfied or totally satisfied with the overall PDS. Table 4.18.--Distribution of satisfaction ratings for the overall PDS. Rating Score N % Totally dissatisfied 1 O 0.0 Dissatisfied 2 17 9.0 Indifferent 3 10 5.3 Satisfied 4 126 66.7 Totally satisfied 5 36 19.0 Missing - 2 -- Relational Findings In this section, the descriptive findings are examined in a relational perspective. That is, satisfaction with overall PDS and each of its components are treated as dependent variables in a series of statistical tests conducted with the supplier variables, the com— pany variables, and the situational variables. Supplier variables.--The variable category of supplier was used as an independent variable in a t-test in which two groups were formed: manufacturers and nonmanufacturers. Table 4.19 sum- marizes the various t-tests performed. In all of the tests for which the t-value was significant at an a level of .10, the respondents were 68 more satisfied with the services provided by the wholesalers or dis- tributors as compared to the manufacturers. One explanation for this finding could be related to the geographical location of the suppliers with respect to their customers: middlemen usually are closer to their customers than are manufacturers. Table 4.19.--Summary of t-tests with supplier category. . Meansa b Dependent Variable Manufac- Middle- t-value p-value turers men Billing procedures 2.16 2.04 1.04 .298 Average delivery time 1.99 1.86 1.00 .318 Delivery time variability 2.26 1.94 2.33 .021 Rush services 2.27 1.79 2.88 .004 Returns policy 2.07 1.91 1.09 .279 Order status information 1.98 1.92 .36 .719 Accuracy in filling orders 1.99 1.82 1.33 .186 Actions on complaints 1.91 1.65 2.30 .022 Order methods 1.90 1.68 2.11 .036 Overall PDS {Question 23) 3.90 4.09 -l.62 .106 Overall PDS weighted ave.) 8.95 8.46 1.26 .211 aVariables billing procedures through order methods were recoded so that the lower the mean, the higher the satisfaction. bDegrees of freedom = 188 except for overall PDS (Question 23) = 186 and for overall PDS (weighted average) = 187. In Table 4.20, a series of Pearson correlation coefficients between the number of supplier variables and satisfaction with overall PDS were calculated. All of the correlation coefficients were very small, but an explanation can be made from the signs of the coefficients 69 that were significant.1 It is apparent that the higher the number of suppliers, the lower the satisfaction with overall PDS. Table 4.20.--Correlations between supplier variables and overall PDS. Overall PDS Supplier Variable Question 23 Wflghted Average r p-value r p-value Number of suppliers used -.02 .401 .06 .223 Number of other suppliers -.15 .030 .14 .038 Number of alternative suppliers -.12 .050 .11 .069 Company variables.--These variables were used to determine if company size had any effect on the satisfaction ratings of the PDS com- ponents and overall PDS. Parametric correlations were run between these variables and overall PDS. None was significantly different from zero. Despite these findings, a further exploration of the rela- tionships between company size and satisfaction with PDS and its com- ponents was conducted via a series of analyses of variance. These ANOVAs were univariate for the overall PDS and multivariate for the PDS components. When the independent variable was company size, measured by sales volume, all the ANOVAs showed no significant differ- ences in the mean satisfaction ratings between the various groups of companies. However, when company size was measured by the number of employees, some significant findings were determined. 1Because of recoding procedures, the two measures of overall PDS should have opposite readings; that is, a lower rating for the weighted average measure means higher satisfaction, and a lower rating for the Question 23 measure means lower satisfaction. 70 For the purpose of this analysis, the variable company size was divided into three groups, according to the number of employees: group l--1ess than 100, group 2--between 100 and 499, and group 3-—500 or more. In the MANOVA, in which the PDS components were the dependent variables, the following hypotheses were tested: Ho: u] = “2 3 that is, the mean of group 1 is H , u f u ’ different from the mean of the l' l 2,3 other two groups Pro: p2 = p3 that is, the mean of group 2 is , different from the mean of group 3 H2. U2 f U3 Table 4.21 presents a summary of the MANOVA performed with company size as the independent variables and the PDS components as dependent variables. Examining the table, one can see that there was no difference between the mean ratings of group 1 and the other two groups combined. But the mean ratings between group 2 and group 3 were significantly different. The variables that were more important in accounting for the differences were returns policy and accuracy in filling orders (at a = .05). For the overall PDS, two ANOVAs were performed to test the same hypotheses detailed above. For the measure of PDS given by the weighted average of PDS components, the groups were not significantly different. However, for the measure of PDS given by Question 23 on the questionnaire, groups 2 and 3 were different but group 1 was not different from the rest. It is interesting that the mean satisfaction ratings for group 2 (medium-size companies) were higher than the mean for group 3 (large companies), indicating that respondents from the 71 larger companies were less satisfied with the overall PDS provided than were respondents from the medium-size companies. Table 4.22 shows a summary of this ANOVA. Table 4.21.--MANOVA: Satisfaction with PDS components by company size (employees). Hypothesis F-valuea p-value H1 (group 1 and groups 2,3) .8938 .5324 H2 (group 2 and group 3) 3.4612 .0007 aThe degrees of freedom were 9 for the numerator and 154 for the denominator in both hypotheses. Table 4.22.--ANOVA: Satisfaction with overall PDS (Question 23) by company size (employees). Hypothesis F-valuea p-value H] (group 1 and groups 2,3) .5299 .4678 H2 (group 2 and group 3) 4.7845 .0302 aThe degrees of freedom were 1 and 162. Situational variables.--The relationships between the situa- tional variables (number of deliveries, backorders, and average order cycle time) and the satisfaction variables are presented in this section. In Table 4.23, parametric correlation coefficients between the situational variables and overall PDS are shown. The interpreta- tions are straightforward: The higher the percentage of backorders, 72 the lower the satisfaction with overall PDS because higher percentages of backorders indicate poor service. Also, the longer the average order cycle time, the lower the satisfaction with overall PDS. Shorter order cycle times indicate better service. Table 4.23.--Correlations between situational variables and overall PDS. Overall PDSa Situational Variable Question 23 Weighted Average r p-value r p-value Number of deliveries -.O9 .154 -.02 .400 Percentage of backorders -.40 .001 .39 .001 Average order cycle time -.28 .001 .15 .023 aBecause of recoding procedures, a lower rating for the weighted average measure means higher satisfaction, and for the Question 23 measure a lower rating means lower satisfaction. Since a considerable number of respondents did not have any backorders (see Table 4.10), the variable backorders was divided into two groups (one with no backorders and the other with backorders). A number of t-tests, summarized in Table 4.24, were conducted with the ratings of satisfaction with PDS and its components. To confirm the findings, the t-tests demonstrated that the groups with no backorders had significantly higher satisfaction ratings than did the group with backorders for all the variables except billing procedures. 73 Table 4.24.--Summary of t-tests with variable percentage backorders. Means Dependent Variablea No t-valueb p-value Backorders Other Billing procedures 2.09 2.22 - .99 .325 Average delivery time 1.67 2.29 -5.36 .000 Delivery time variability 1.82 2.51 -5.68 .000 Rush services 1.78 2.55 -4.91 .000 Returns policy 1.89 2.23 -2.30 .023 Order status information 1.74 2.23 -3.59 .000 Accuracy in filling orders 1.65 2.30 -5.74 .000 Actions on complaints 1.66 2.08 -3.70 .000 Order methods 1.75 1.94 -1.80 .074 Overall PDS (Question 23) 4.19 3.69 3.45 .001 Overall PDS (weighted ave.) 7.99 9.78 -4.85 .000 aPDS components variables were recoded so that the lower the mean, the higher the satisfaction. 170 bDegrees of freedom = 174 except for overall PDS (Question 23) = Importance of PDS Components The findings concerning the importance of PDS components are presented in the same manner as the satisfaction findings; that is, the descriptive findings are discussed first, followed by the rela- tional findings. Descriptive Findings, The mean ratings of the importance of the PDS components, as perceived by the respondents, are presented in Table 4.25. To allow for a rank-ordering of the importance of the PDS components, a Wilcoxon matched-pairs signed test was conducted for every difference between mean rankings (see Table 4.26). Using the data shown in Table 4.25 74 and the results of the various Wilcoxon tests, the following rank— order of the importance of PDS components was obtained: 1. Accuracy in filling orders--regarded as the most important PDS component 2. Average delivery time Rush services and billing procedures Actions on complaints Order status information Delivery time variability \IOTU'l-b Returns policy and order methods--considered the least important PDS components Table 4.25.--Importance ratings for PDS components. Standard PDS Components Mean Deviation Billing procedures 3.932 1.917 Average delivery time 3.126 1.802 Delivery time variability 5.262 1.499 Rush services 3.864 1.813 Returns policy 5.623 1.083 Order status information 4.822 1.497 Accuracy in filling orders 2.267 1.672 Actions on complaints 4.576 1.550 Order methods 5.686 1.375 Relational Findings The importance rankings of the PDS components were treated as dependent variables to permit the drawing of possible relationships with the supplier variables and the company variables. 75 Table 4.26.--Summary of the Wilcoxon test for mean ratings of impor- tance for PDS components. Pair of Variables z-valuea p-value Billing procedures/average delivery time -3.981 .000 Billing procedures/delivery time variability 6.102 .000 Billing procedures/rush services -O.3ll .377 Billing procedures/returns policy 8.729 .000 Billing procedures/order status information 4.530 .000 Billing procedures/accuracy in filling orders -6.970 .000 Billing procedures/actions on complaints 3.446 .000 Billing procedures/order methods 8.414 .000 Average delivery time/delivery time variability 9.181 .000 Average delivery time/rush services 3.862 .000 Average delivery time/returns policy 10.269 .000 Average delivery time/order status information 7.795 .000 Average delivery time/accuracy in filling orders -4.064 .000 Average delivery time/actions on complaints 6.577 .000 Average delivery time/order methods 10.196 .000 Delivery time variability/rush services -6.897 .000 Delivery time variability/returns policy 2.932 .001 Delivery time variability/order status information -2.903 .001 Delizgry time variability/accuracy in filling _]0.775 .000 Delivery time variability/actions on complaints -3.836 .000 Delivery time variability/order methods 2.724 .003 Rush services/returns policy 8.704 .000 Rush services/order status information 4.840 .000 Rush services/accuracy in filling orders -7.341 .000 Rush services/actions on complaints 3.934 .000 Rush services/order methods 8.381 .000 Returns policy/order status information -5.803 .000 Returns policy/accuracy in filling orders —lO.90l .OOO Returns policy/actions on complaints -7.108 .000 Returns policy/order methods 0.669 .251 Orgegesgatus information/accuracy in filling _10.399 .000 Order status information/actions on complaints 1.648 .049 Order status information/order methods 5.579 .000 Accuracy in filling orders/actions on complaints 10.188 .000 Accuracy in filling orders/order methods 10.964 .000 Actions on complaints/order methods 6.924 .000 a2 = T - “T where T = the smaller sum of the like-signed ranked. 0 T 76 Supplier variables.--The variable category of supplier was used as an independent variable in a t-test with two groups: manufacturer and nonmanufacturer. Table 4.27 gives a summary of the t-tests performed. At an a level of .05, only two differences were ' significant: order status information and actions on complaints. The respondents for whom the supplier was a manufacturer ranked both com- ponents as more important than did the respondents with middlemen as suppliers. But at an a level of .10, two more differences were signifi- cant: billing procedures and rush services. However, the respondents for whom the supplier was a manufacturer ranked both of these components as less important than did the respondents with middlemen as suppliers. Table 4.27.--Summary of t-tests with supplier category. Means of Categories Dependent Variable Manufac- Middle- t-valuea p-value turers men Billing procedures 4.13 3.64 1.69 .093 Average delivery time 3.04 3.27 -.84 .401 Delivery time variability 5.23 5.27 -.17 .866 Rush services 4.01 3.56 1.63 .105 Returns policy 5.67 5.53 .84 .402 Order status information 4.50 5.27 —3.38 .001 Accuracy in filling orders 2.15 2.42 -1.11 .270 Actions on complaints 4.40 4.94 -2.32 .022 Order methods 5.80 5.62 .87 .384 aDegrees of freedom = 188. 77 Company_variables.—-Of this group of variables, the ones related to company size were selected for exploring relationships with the importance of PDS components. A series of parametric Pearson correlation coefficients between company-size variables and the importance of P05 components is presented in Table 4.28. All of the coefficients were very small, indicating a low degree of associa- tion. However, the sign of the coefficient, provided it was signifi- cant, gave some insights into the direction of the relationship. Therefore, statements like the following can be made (at an a level of .10): 1. The larger the company (in terms of number of employees), the higher the importance of delivery time variability, rush services, accuracy in filling orders, and actions on complaints; and the lower the importance of billing procedures and order methods. 2. The larger the company (in terms of sales volume), the higher the importance of delivery time variability, rush services, and accuracy in filling orders; and the lower the importance of bill- ing procedures. Note that some relationships (delivery time variability, rush services, accuracy in filling orders, and billing procedures) occurred in both company-size variables. To explore further the relationships between company variables land the importance rankings of the PDS components, two sets of ANOVAs inere performed: one univariate with each component of PDS as dependent variable and one multivariate with the PDS components as dependent variables. I 78 Table 4.28.--Correlations between company variables and importance of PDS components. Company Size Dependent Variable # of Employees Sales Volume r p-value r p-value Billing procedures .26 .001 .25 .001 Average delivery time ,0] .47] -,03 .14] Delivery time variability -,]3 .035 -,15 .025 Rush services -.14 .031 -.10 .081 Returns policy .02 .366 .02 .413 Order status information -.01 .433 -.O4 .291 Accuracy in filling orders —.18 .007 -.15 .021 Actions on complaints -.12 .052 .06 .213 Order methods .16 .014 .01 .469 In Table 4.29 the univariate ANOVA between company size, measured in terms of number of employees, and the importance of P05 components is shown. The significant differences (atcx==.10) among the groups occurred in the variables billing procedures, delivery time variability, and accuracy in filling orders. For the billing procedures, group 4 (companies with 1,000 or more employees) had the lowest importance ranking. For delivery time variability, the group of large companies (group 4) had the highest importance ranking. However, for accuracy in filling orders, the group of smaller companies (group 1, with less than 100 employees) had the lowest importance rank- ing. Table 4.30 shows a summary of the preceding interpretations. Note that a significant F-value assures that at least two groups were different. 79 Table 4.29.--Summary of ANOVAs between company sizea and importance of PDS components. Dependent variable MS Between MSE F-valueb p-value Billing procedures 18.130 3.508 5.168 .002 Average delivery time 2.543 3.283 .775 .510 Delivery time variability 4.824 2.211 2.182 .092 Rush services 4.993 3.259 1.532 .208 Returns policy .026 1.191 .022 .996 Order status information 1.219 2.386 .511 .675 Accuracy in filling orders 12.816 2.644 4.847 .003 Action on complaints 3.599 2.405 1.497 .217 Order methods 3.057 1.781 1.716 .165 aThe variable company size was divided into four groups with respect to number of employees: l--1ess than 100, 2--lOO to 499, 3--500 to 999, and 4--more than 1,000. bThe degrees of freedom were 3 and 187. Table 4.30.--Summary of the mean rankings of importance for different groups of company size (in terms of number of employees). Mean Ranking of Importance Groups of . . . . . Company Size Billing Delivery Time Accuracy in Procedures Variability Filling Orders l--Less than 100 3.27 5.25 2.90 2--100 to 499 4.06 5.45 1.91 3--500 to 999 4.12 5.12 2.44 4--l,OOO or more 4.96 4.60 1.92 The univariate ANOVAs between company size, measured in terms of sales volume, and importance of PDS components are presented in Table 4.31. The company-size variable was divided into five groups, with the same intervals as in the original data-collection instrument. That is, group 1--20 to 49, group 2--50 to 99, group 3—-100 to 499, group 4--500 to 999, and group 5--l,000 or more (values in millions of 80 cruzeiros). As can be seen from the table, the groups were signifi- cantly different (at a = .10) on the following components: billing procedures, delivery time variability, accuracy in filling orders, and actions on complaints. Table 4.31.--Summary of ANOVAs between company sizea and importance of PDS components. Dependent Variable MS Between MSE F-valueb p-value Billing procedures 12.525 3.668 3.414 .010 Average delivery time 2.969 3.247 .915 .457 Delivery time variability 4.659 2.264 2.058 .088 Rush services 3.271 3.329 .982 .418 Returns policy 1.406 1.207 1.165 .328 Order status information 1.976 2.375 .832 .506 Accuracy in filling orders 7.963 2.753 2.893 .024 Action on complaints 7.099 2.339 3.053 .019 Order methods .378 1.818 .208 .934 a . . Company Size measured in terms of sales volume. bDegrees of freedom were 4 and 179. To allow for further interpretations of these findings, the means of the importance rankings of PDS components by each one of the company-size groups are presented in Table 4.32. One pattern that can be observed is that the larger the company, the lower the impor- tance of billing procedures. For delivery time variability, the group of larger companies (1,000 or more millions of cruzeiros) had the highest importance rating. For the other two components, accuracy in filling orders and actions on complaints, even though at least two groups were different, there seemed not to be an apparent relationship 81 pattern. However, for both components, group 4 (500 to 999 millions of cruzeiros) showed the highest mean importance rankings. Table 4.32.--Summary of the mean rankings of importance for different company-size groups (in terms of sales volume). Mean Ranking of Importance Company-Size Groupsa Billing 09%};gry ificgiiiifig Actions on Procedures Variability Orders Complaints l-— 20 to 49 3.11 5.26 2.63 4.44 2-- 50 to 99 3.32 5.38 3.03 4.68 3--1OO to 499 4.17 5.43 2.08 4.58 4--500 to 999 4.45 5.15 1.65 3.65 5--1,000 or more 4.53 4.53 2.23 5.17 a . . . . Values are in millions of cruzeiros. In conclusion, two MANOVAs were conducted with company size as independent variable and the importance ranking of POS components as dependent variables. In the first MANOVA, which is presented in Table 4.33, the company-size variable was divided into three groups according to the number of employees: group l--less than 100, group 2--between 100 and 499, and group 3--500 or more. The follow- ing hypotheses were tested: Ho: u1 = 112,3 that is, the mean of group 1 is H , f different from the mean of the 1' pl u2,3 other two groups “If 112 = u3 that is, the mean of group 2 is different from the mean of group 3 82 Table 4.33.--MANOVA: Importance of PDS components by company size (employees). Hypothesis F-valuea p-value H] (group 1 and groups 2,3) 2.3143 .0181 H2 (group 2 and group 3) 1.9735 .0459 aDegrees of freedom were 9 and 154. The two null hypotheses were rejected; in fact, there were differences among the three groups with respect to their ratings of the importance of PDS components. In the case of H], the components billing procedures and accuracy in filling orders were the most impor- tant in accounting for the differences in the means (at a = .05). For Hz, the variables billing procedures and average delivery time were the most important, also at a = .05. For the second MANOVA, the variable company size, measured in terms of sales volume (given in millions of cruzeiros), was grouped in the following manner: l--less than 100, 2--lOO to 499, and 3--500 or more. The hypotheses to be tested were the same as the ones out- lined for the company-size variable measured in terms of number of employees. The summary of this MANOVA is given in Table 4.34. The difference between group 1 and the other groups was significant, and the variables billing procedures and accuracy in filling orders were the nest important in determining the difference. However, groups 2 .and 3 were not statistically different, and therefore the null hypothe- sis could not be rejected. 83 Table 4.34.--MANOVA: Importance of PDS components by company size (sales). Hypothesis F-valuea p-value H] (group 1 and groups 2,3) 2.2745 .0204 H2 (group 2 and group 3) 1.1780 .3132 aThe degrees of freedom were 9 and 154. PDS and Other Factors Affecting the Patronage Decision In this section, the findings related to the factors affect- ing the patronage decision are discussed. As in the previous sections, the analysis is divided into two parts: the first, descriptive, and the second, relational. Descriptive Findings The mean ratings of the importance of a group of factors in selecting suppliers, as perceived by the respondents, are given in Table 4.35. The same procedure for comparing the mean rankings (Wilcoxon signed test) used earlier for ranking the importance of PDS components was used with the purchasing factors. As can be seen in Table 4.36, all of the differences were significant, enabling the presentation of the following rank of factors, in descending order of 'hnportance, affecting the selection of suppliers: 1. Product quality 2. Price PDS 3 4. Geographical location of the supplier 84 5. Required minimum order size 6. Supplier management 7. Reciprocity Table 4.35.--Importance ratings of purchasing factors. - Standard PurchaSing Factors Mean Deviation Product quality 1.379 .956 PDS 3.421 1.265 Price 2.116 .877 Supplier management 5.354 1.409 Geographical location of supplier 4.484 1.483 Required minimum order size 5.100 1.371 Reciprocity 5.751 1.375 Relational Findings In this section, the rankings of the importance of the pur- chasing factors are examined in a relational perspective. These rankings were dependent variables in a series of statistical tests in which supplier and company variables were treated as independent variables. Supplier variable.--The supplier-variable category was con- sidered as an independent variable in a series of t-tests with two groups: manufacturers and nonmanufacturers. Table 4.37 presents a summary of all the tests that were conducted. The differences were significant (at a = .10) for the first four factors listed in the table. The respondents for whom the suppliers were manufacturers 85 Table 4.36.--Summary of the Wilcoxon test for mean ratings of the importance of the purchasing factors. Pair of Variables z-valuea p-value Product quality/PDS 10.410 .000 Product quality/price 6.966 .000 Product quality/supplier management 11.533 .000 Product quality/geographical location of supplier 11.322 .000 Product quality/required minimum order size 11.551 .000 Product quality/reciprocity 11.491 .000 PDS/price -8.629 .000 PDS/supplier management 9.884 .000 PDS/geographical location of supplier 5.835 .000 PDS/required minimum order size 8.734 .000 PDS/reciprocity 10.329 .000 Price/supplier management 11.249 .000 Price/geographical location of supplier 11.129 .000 Price/required minimum order size 11.497 .000 Price/reciprocity 11.538 .000 Supplier management/geographical location of _ supplier 5.057 .000 Supplier management/required minimum order _] 644 050 size ' ' Supplier management/reciprocity 2.733 .003 Geographical location of supplier/required minimum order size 3'640 '000 Geographical location of supplier/reciprocity 6.732 .000 Required minimum order size/reciprocity 4.565 .000 a2 = T ' L1T where T = the smaller sum of the like-signed ranked. OT 86 ranked both product quality and P05 lower (that is, more important) than did the respondents who used middlemen as their suppliers. An inverse relationship (that is, less important) was observed for the factors price and supplier management. Table 4.37.--Summary of t-tests with supplier category. Means of Categories . . a Dependent Variable Manufac- Middle- t-value p-value turers men Product quality 1.26 1.61 -2.42 .016 PDS 3.31 3.65 -l.81 .072 Price 2.20 1.94 1.98 .049 Supplier management 5.49 5.11 1.79 .076 Geographical location of supplier 4.52 4.41 .51 .612 Required minimum order size 5.10 5.06 .17 .863 Reciprocity 5.85 5.53 1.44 .150 aThe degrees of freedom were 188. Company variables.--The company-size variables, measured in terms of sales volume and number of employees, were correlated with each of the factors in selecting suppliers to explore the degree of association between them. Table 4.38 presents a summary of the parametric Pearson correlation coefficients. Even though the coefficients were small, the signs of the significant associations (at a.= .05) were relevant, and the following statements can be made: 'The larger the company, measured either by sales volume or by number of employees, the higher the importance of product quality and the lower'the importance of price as factors in selecting suppliers; 87 and the larger the company, measured in terms of number of employees, the higher the importance of PDS in selecting suppliers. Table 4.38.--Correlations between company variables and factors in selecting suppliers. Company Size Factors in Selecting Suppliers # of Employees 53195 Volume r P-value r p-value Product quality -.25 .001 -.21 .003 PDS -.12 .047 -.08 .151 Price .12 .050 .21 .002 Supplier management .10 .091 .03 .340 Geographical location of supplier .08 .145 -.04 .306 Required minimum order size .03 .366 .08 .135 Reciprocity .11 .067 .02 .387 Two MANOVAs were used to examine further the differences in the ranking of importance of factors in selecting suppliers with respect to company size. In the first MANOVA, presented in Table 4.39, the company-size variable was divided into three groups according to the number of employees: group l--1ess than 100, group 2--between 100 and 499, and group 3--500 or more. The following hypotheses were tested: H0: u] = pz 3 that is, the mean of group 1 is ’ different from the mean of the H]: u] = u2,3 other two groups ”if ‘02 = p3 that is, the mean of group 2 is different from the mean of group 3 88 At an a value of .10, the two null hypotheses were rejected; there were differences among the groups with respect to the rankings in the set of dependent variables. For the difference between group 1 and the other two groups, product quality was the most important deter- minant of the difference. And for the difference between group 2 and group 3, PDS was the most important factor in accounting for the difference. Table 4.39.--MANOVA: Importance of purchasing factors by company size (employees). Hypothesis F-valuea p-value H1 (1 and 2,3) 1.7663 .0978 H2 (2 and 3) 1.7541 .1004 aThe degrees of freedom were 7 and 156. For the second MANOVA, the variable company size, measured in terms of sales volume (given in millions of cruzeiros), was grouped as follows: l--1ess than 100, 2--lOO to 499, and 3—-SOO or more. The hypotheses tested were the same ones as cited previously for the company-size variable measured in terms of number of employees. From Table 4.40, it can be seen that only the difference between group 1 and the other two groups was significant at an a value of .10. For this difference, price was the major determinant. 89 Table 4.40.--MANOVA: Importance of purchasing factors by company size (sales). Hypothesis F-valuea p-value H] (group 1 and groups 2,3) 1.7481 .1020 H2 (group 2 and group 3) 1.4599 .1858 aThe degrees of freedom were 7 and 156. P05 and the Patronsge Decision In this section, the data are analyzed to provide insights for examining the hypotheses that were instrumental in fulfilling one of the objectives of this research--that is, to explore relationships between situational variables of the buying process, supplier vari- ables, company variables, and satisfaction variables with the per- ceived importance of P05 in selecting suppliers. Situational Variables Hypothesis 1: The greater the number of deliveries, the greater the importance of PDS in selecting suppliers. The statistical hypotheses derived from this substantive hypothesis are: Ho: r z 0 where r = Pearson correlation coefficient between number of deliveries H]: r < O and importance of PDS The correlation coefficient between the two variables was -.06 with a p-value of .252. Thus, the null hypothesis could not be rejected. 90 To examine further the relationship formulated in the hypothe- sis, the variable number of deliveries was divided into three groups according to the frequency of deliveries per month: group 1--l or less, group 2--more than 1 and 3 or less, and group 3--more than 3. A univariate ANOVA with number of deliveries as the independent vari- able and importance of PDS as the dependent variable was performed. The F-value was .540 with df==2, 142 and p-value of .584. Therefore, there seemed to be no difference among the three groups. Hypothesis 2: The higher the proportion of backorders, the greater the importance of PDS in selecting suppliers. The testable hypotheses were similar to the preceding ones; that is: H : r 2 0 where r = Pearson correlation coefficient 0 between percentage of backorders H]: r < O and importance of PDS The coefficient was zero with a p-value of .481. Therefore, the null hypothesis could not be rejected. Since the number of respondents who had no backorders was fairly large (96--see Table 4.11), this variable was divided into two groups (no backorders and backorders) to allow for a t-test to examine the differences between these groups. The t-value was .35 with a p-value of .474. Thus the hypothesis of no difference could not be rejected. Hypothesis 3: The greater the average order cycle time, the greater the importance of PDS in selecting suppliers. The following hypotheses were tested: 91 Ho: r 3 0 where r = Pearson correlation coefficient between average order cycle time H]: r < 0 and importance of PDS The coefficient of correlation was -.13 with a p-value of .041. Even though the absolute value of the coefficient was fairly small, the sign of the coefficient implied that the direction of the association was negative; therefore, the null hypothesis was rejected. To examine this relationship further, a breakdown of the means of the average order cycle time by every rank of importance of PDS is presented in Table 4.41. The figures in the table suggest that the respondents with low ranks of importance (6 and 7) did, in fact, have lower average order cycle times than did the respondents with high ranks of importance (1 and 2), but the middle groups (3, 4, and 5) were not very different from the high-importance groups. Therefore,ii:seems that only substantial changes in order cycle time would affect the importance ranking of PDS. Table 4.41.--Breakdown of average order cycle time by importance of PDS. Mean of Average N Rank of Importance Order Cycle Time 1 26.0 6 2 30.2 26 3 22.3 72 4 25.1 32 5 21.6 18 6 13.8 12 7 10.5 2 Supplier Variables Hypothesis 4: The greater the number of alternative suppliers available, the lower the importance of PDS in selecting suppliers. 92 The statistical hypotheses derived from the substantive hypotheses are the following: H0: r 5 0 where r = Pearson correlation coefficient between the number of alternative H]: r > O suppliers and importance of PDS The coefficient of correlation was .06 with a p-value of .222. Therefore, the null hypothesis could not be rejected. The association between the importance of P05 as a factor in selecting suppliers and the other two supplier variables (specifically, number of suppliers inalized and the number of other suppliers) was also explored. The Pearson correlation coefficients are pre- sented in Table 4.42. Interpreting the data in this table, the fol— lowing statement can be made (at a = .10): The higher the number of suppliers utilized, the lower the importance of PDS in selecting suppliers . Table 4.42.--Spearman correlation coefficients. Variable Pair r p-value Suppliers utilized/importance of P05 .12 .060 Other suppliers/importance of P05 .01 .475 Company Variables Hypothesis 5: The larger the company, either in terms of number of employees or in sales volume, the greater the importance of PDS as a factor in selecting suppliers. From this hypothesis, two sets of testable hypotheses can be derived: 93 Pearson correlation coefficient between company size, measured in H]: r < 0 terms of number of employees, and importance of PDS H : r 3 0 where r Pearson correlation coefficient between company size, measured in H']: r' < 0 terms of sales volume, and impor- tance of PDS H' : r' 3 0 where r The correlation coefficients were the following: for company size (number of employees) and importance of PDS, -.12 with p-value of .047; for company size (sales volume) and importance of PDS, -.08 with p-value of .151. Therefore, the null hypothesis H0 was rejected, and the null hypothesis H2) could not be rejected. Two ANOVAs were performed with the company-size variables as independent variables and the importance of PDS in selecting suppliers as the dependent variable. Each ANOVA is discussed and analyzed separately. In the first ANOVA, the company-size variable was divided into four groups according to the number of employees: 1--1ess than 100, 2--lOO to 499, 3--5OO to 999, and 4--more than 1,000. In the second, the company-size variable, measured by sales volume (given in millions of cruzeiros), was divided into the following groups: 1--from 20 to 49, 2--from 50 to 99, 3--from 100 to 499, 4--from 500 to 999, and 5--1,000 or more. A summary of the two ANOVAs is presented in Table 4.43. At an a level of .10, both ANOVAs were significant, meaning that at least two of the groups were different. 94 Table 4.43.--Summary of ANOVAs between company size and importance of PDS. Group MS Between df MSE df F-value p-value Number of employees 3.810 3 1.569 184 2.429 .067 Sales volume 3.411 4 1.573 176 2.168 .074 To examine further the differences in mean rankings of impor- tance of PDS in selecting suppliers within the two company-size groups, Tables 4.44 and 4.45 are presented. As shown in the first table, respondents from the two groups of larger companies (groups 3 and 4) tended to rank importance of PDS higher than did those from the groups of smaller companies (groups 1 and 2). In fact, the major difference seemed to occur between groups 2 and 3. In the second table, groups 2, 4, and 5 were probably different than groups 1 and 3. In fact, groups 3 and 5 were different, with respondents from the larger-companies group (group 5) ranking importance of PDS higher than did respondents from smaller companies (group 3). Table 4.44.--Mean rankings of importance of PDS by company size (number of employees). Group (number of employees) Mean l--1ess than 100 3.33 2-—1OO to 499 3.57 3--500 to 999 2.75 4--l,OOO or more 3.16 95 Table 4.45.--Mean rankings of importance of P05 by company size (sales volume). Group (sales volumea) Mean l-- 20 to 49 3.56 2-- 50 to 99 3.14 3--1OO to 499 3.71 4--500 to 999 3.25 5--1,000 or more 3.07 a . . . . . Given in millions of cruzeiros. Satisfaction Variables Hypothesis 6: The greater the satisfaction with PDS, the lower its importance as a purchasing factor. Since there were two measures of satisfaction with PDS (one given by the respondents [Question 23] and the other computed by the researcher [weighted average of PDS componentSJ).two sets of testable hypotheses could be derived from the substantive hypothesis: Ho: r 3 0 where r = Pearson correlation coefficient between satisfaction with PDS H]: r < 0 (Question 23) and importance of PDS FPO: r' 5 0 where r = Pearson correlation coefficient between satisfaction with PDS H']: r' > 0 (weighted average) and importance of PDS The coefficients of correlation were r = .03 with p-value (if .338 and r' = —.02 with p-value of .372. Therefore, neither null hypothesis could be rejected. Every one of the satisfaction ratings of PDS components was cxarrelated with the importance of P05 in selecting suppliers. Only 96 the Pearson correlation coefficient between billing procedures and importance of PDS was significant at a .10: r = -.13 with p-value== .036. Thus it can be said that the higher the satisfaction with bill- ing procedures, the lower the importance of PDS. Comparative Analysis In this section a comparative analysis of the findings of this research and those reported by Perreault and Russ (1976a) is presented. The analysis is divided into three major areas. The first deals with satisfaction with PDS and its components. The second concerns the importance of particular factors in selecting suppliers. The last examines the relationships between situational variables (deliveries, backorders, and order cycle time), a supplier variable (alternative suppliers), and a satisfaction variable (overall PDS) with the importance of P05 in selecting suppliers. The emphasis of the analysis is descriptive because the conclusions and implica- tions of the comparative findings are explored in Chapter V. Satifaction With PDS and Its Components In Table 4.46, the satisfaction ratings of PDS components are outlined with respect to the rank order of satisfaction and the mean 1 ratings of satisfaction. The following are some highlights of that table: 1See Chapter III, p. 46, for an explanation of the scaling and measurement procedures of both studies. 97 --in both studies, delivery time variability was the component with which the respondents were least satisfied; however, in this research, for average delivery time, the respondents were not as dis- satisfied as were the respondents in the Perreault and Russ study; --accuracy in filling orders and order methods were components with which respondents were highly satisfied; --respondents in the Perreault and Russ study were most satis- fied with the component billing procedures, but respondents in the present study were next-to-least satisfied with that component. Table 4.46.--Satisfaction ratings of P05 components. This Research Perreault 8 Russ PDS Component Studya Rank Mean Rank Mean Billing procedures 8 3.84 1 1.99 Average delivery time 4 4.02 8 2.94 Delivery time variability 9 3.81 9 3.15 Rush services 7 3.86 6 2.68 Returns policy 6 3.90 3 2.21 Order status information 5 3.98 7 2.90 Accuracy in filling orders 3 4.06 2 2.00 Actions on complaints 2 4.12 5 2.35 Order methods 1 4.13 3 2.21 aSource: Perreault and Russ (1976a, p. 8). To determine what components of PDS are most likely to increase customer satisfaction, a series of correlations between satis- faction ratings of every component and satisfaction with overall PDS are presented in Table 4.47. In both studies, returns policy had the lowest correlation with satisfaction with overall PDS. Billing 98 procedures also had low correlations, whereas delivery time varia- bility presented high correlation coefficients. The only discrepancy between the two studies seemed to be with respect to accuracy in filling orders: in this research it showed the highest correlation with satisfaction with overall PDS, whereas in the Perreault and Russ study it had one of the lowest correlations. Table 4.47.--Correlations between components and overall PDS. Overall PDS Component This Research Perreggl§y& Russ Billing procedures .48 .39 Average delivery time .59 .76 Delivery time variability .63 .72 Rush services .61 .59 Returns policy .34 .44 Order status information .53 .67 Accuracy in filling orders .63 .46 Actions on complaints .56 .56 Order methods .37 .56 aSource: Perreault and Russ (1976a, p. 8). Another comparison was made regarding the satisfaction with PDsiand the expectations raised by suppliers with their customers by seeking feedback about the services provided. Table 4.48 summarizes the findings concerning such feedback; this information was obtained by asking the following question (number 29 on the questionnaire): "In: your suppliers of this product check with you to see if the ser— vices they are providing are adequate in meeting your needs?" These firudings were identical, in degree, to the ones presented by Perreault 99 and Russ (1976a, p. 9). Examining the data in the table, one can affirm that the respondents who checked response 3 to this question were the least satisfied with the services provided. Also, the respondents who were not contacted by their suppliers and did not see any need for such a contact (response 5) were highly satisfied with the services they were receiving. Table 4.48.--Supplier sensitivity to purchasers' service needs and purchaser satisfaction with PDS.a Percent Mean Rating of Responding Overall PDSb Response This Perreault This Perreault Study and Russ Study and Russ 1. Yes, but there are no needed improvements '26 '18 4'2] 1°62 2. Yes, and they have made .49 .50 4.08 2.21 improvements 3. Yes, but the services .06 .14 2.63 3.3] did not improve 4. No, but I would like them to do so .13 .14 3.43 2.87 5. No, and it is not necessary .06 .04 4.41 1.33 aTitle taken from Perreault and Russ (19766, P- 9). bIn this study a high mean rating means highly satisfied, whereas iii the Perreault and Russ study it is the oppOSite. Importance of Factors iii Selecting Suppliers In Table 4.49, a comparison of the rankings of the importance of: different factors in selecting suppliers is exhibited. The findings 100 of both studies were similar, and some relevant comments are mentioned below: --product quality had the highest ranking (that is, most important) in both studies; --in the Perreault and Russ study, PDS was second only to product quality, but in this research price was regarded as more important than PDS; a possible explanation for this difference is the prevailing situation in the Brazilian economy during 1980. The real interest rates were negative; thus companies were hedging against inflation by building up inventories.1 Therefore, price was an important factor; --reciprocity was the least important factor in both studies. Table 4.49.--Ranks of importance of purchasing factors. Importance Rankings Factor This Research Perreault & Russ Studya Product quality 1 l PDS 3 2 Price 2 3 Supplier management 6 4 Distance to supplier 4 5 Required order size 5 6 Minority/small business n.a 7 Reciprocity 7 8 aSource: Perreault and Russ (1976a, p. 5). 1In a forthcoming research report by Luce et al., that situa- tion was found to be true; the average price of raw materials for both mechanical and metallurgical industries rose 130% during 1980, whereas the average interest rate prevailing in the market for the same period was between 80% and 85% a year, with the inflation rate for 1980 at about 110%. 101 Relationships The framework for this part of the comparative analysis was an examination of the common hypotheses involving the importance of PDS as a factor in selecting suppliers. Before proceeding to the analysis, a note on the comparability of the relationships is pre- sented. The data obtained in the two studies with respect to the situational variables of the buying process (deliveries, backorders, and order cycle time) were different: In this research the responses were product and supplier specific (that is, the respondents considered only the most important supplier of the chosen product), whereas in the Perreault and Russ study they were only product specific. Thus, some of the comparisons should be made only in terms of degree. According to what was reported by Perreault (1973), no statis- tical significance was found in any of the hypothesized relationships between situational variables (for Perreault they included number of 1 order cycle time, number of alter- deliveries, backorders cancelled, native suppliers, and satisfaction with overall PDS) and importance of PDS in selecting suppliers. In this research, only one of the com- parable hypothesized relationships2 was significant: order cycle time and importance of P05; that is, the greater the order cycle time, the greater the importance of P05 in selecting suppliers. Therefore, the 1See footnote on page 43, Chapter III. 2Another relationship was significant (company size [by number of employees] and importance of PDS) but could not be compared with the U.S. findings because it was not researched. 102 outcomes of both studies were very similar with respect to the hypothesized relationships. To examine further these relationships, Perreault and Russ grouped the respondents "according to their views on the importance of physical distribution service and the average response on each situational variable was computed for each group" (1976a, p. 7). Groups with a low number of respondents (n < 6) were not considered. In this research the same procedure was followed so that comparisons between the two studies could be made. Table 4.50 summarizes this procedure, and an analysis of the findings follows. Relationship 1: The greater the number of deliveries, the greater the importance of PDS. The findings were consistent with the relationship hypothesized in the Perreault and Russ study: "the importance category . . . of PDS increases monotonically with the number of deliveries" (1976a, p. 7). However, in the present research, this pattern was not observed. The only difference seemed to exist between the lowest importance category and the others (recall that an ANOVA was performed with different groups of frequencies of deliveries per month and the importance of PDS, and it was not significantl). Relationship 2: The greater the number of alternative suppliers available, the lower the importance of PDS. In both studies the relationship was not observed, but two major differences were found: (1) in the Perreault and Russ study, the number of "suppliers differs substantially by importance category" 1See page 90, Chapter IV. 103 (1976a, p. 7), and in this study only one group seemed to be different; and (2) the lowest importance group in this research had the highest number of alternative suppliers, whereas in the Perreault and Russ study the lowest importance group had the lowest number of suppliers. Table 4.50.--Mean responses on variables by importance of PDS. Importance Levels of PDSa Variables (Less (More Important) Important) Number of deliveries Perreault and Russb n.a. n.a. 52.3 107.4 174.0 414.0 This study 23.1 33.9 39.0 34.0 35.9 32.3 Number of alternative suppliers Perreault and Russ n.a. n.a. 11.8 29.5 21.6 17.6 This study 13.7 7.3 8.0 9.7 7.5 9.0 Percentage of backorders Perreault and Russ n.a. n.a. 21.4 19.2 22.6 18.0 This study 5.5 6.9 4.6 5.4 6.7 2.7 Satisfaction with PDS Perreault and Russ n.a. n.a. 3.3 2.6 2.5 2.1 This study 4.1 4.0 3.9 4.0 3.9 3.8 Average order cycle time Perreault and Russ n.a. n.a. 39.6 24.9 40.5 32.9 This study 13.8 21.6 25.1 22.3 30.2 26.0 aSome response categories had too few respondents to provide mean- ingful averages and are identified by the "n.a." entry. bA11 data on the Perreault and Russ study were taken from Perreault and Russ (1976a, p. 7). Relationship 3: The greater the satisfaction with PDS, the lower its importance in the evalua- tion. 104 In the Perreault and Russ study, "as the importance of physical distribution service increases, satisfaction increases“ (1976a, p. 8). On the other hand, in this study satisfaction with PDS did not affect the ranking of its importance. Relationship 4: The greater the average delivery time, the greater the importance of P05. "No apparent relationship is observed between the importance of physical distribution service and average delivery time" in the Perreault and Russ study (1976a, p. 8). In this research, the rela- tionship was found to be significant, but it seems that only consider- able reductions in order cycle time could lower the perceived importance of PDS. Relationship 5: The higher the proportion of back— orders, the greater the importance of P05.1 In neither study was the relationship observed, and the per- centage of backorders was "approximately the same across the impor- tance levels" in the Perreault and Russ study (1976a, p. 7). Two other meaningful comparisons can be made by further exam- ining the data presented in Table 4.50. The respondents in this study had, for every importance category, a smaller number of alternative suppliers than did the respondents in the Perreault and Russ study. But they had, in every importance category, a shorter average order cycle time than did the subjects in the Perreault and Russ research. 1This relationship was not hypothesized by Perreault and Russ, taut since the data were available the comparisons were made possible. 105 Summary of Findings The research objectives and hypotheses, with the findings related to each one, are summarized in this final section of Chapter IV. Objective 1: Satisfaction To investigate the perceived satisfaction of buyers with overall PDS and with each of its components. The mean satisfaction rating of the overall PDS (measured on a five-point scale ranging from totally dissatisfied to totally sat- isfied) was 3.96. The majority of the respondents (85.7%) were either "satisfied" or "totally satisfied" with the service they were receiving from their suppliers. With respect to the specific PDS components, the respondents were most satisfied with order methods (mean: 4.134) and action on complaints (mean: 4.124), and they were least satisfied with rush services (mean: 3.586), billing procedures (mean: 3.842), and delivery time variability (mean: 3.807). Objective 2: Importance To rank, in order of importance, different components of PDS as perceived by purchasers. The component with the highest importance ranking was accuracy in filling orders, with a mean ranking of importance of 2.28 (l = most important); the components returns policy and order methods had the lowest importance rankings, with means of 5.62 and 5.69, respectively. Objective 3: Importance To compare the importance of PDS with other factors influencing patronage decisions. 106 Product quality was regarded as the most important factor in selecting suppliers, whereas price was second and PDS third. Objective 4: Relationships To explore relationships between situational variables of the buying process, supplier vari- ables, company variables, and satisfaction vari- ables with the perceived importance of P05 in selecting suppliers. This objective was operationalized in the form of hypotheses. Each of the hypotheses is repeated below, with the outcome of the test for every null hypothesis derived from each of the substantive hypothe- ses. Hypothesis 1: The greater the number of deliveries, the greater the importance of PDS in selecting suppliers. Hypothesis 2: The higher the proportion of backorders, the greater the importance of PDS in selecting suppliers. These two null hypotheses were not rejected. Moreover, no other relationship between each pair of variables was found to be significant. Hypothesis 3: The greater the average order cycle time, the greater the importance of PDS in selecting suppliers. The null hypothesis was rejected and the hypothesized rela- tionship between the variables confirmed. Hypothesis 4: The greater the number of alternative suppliers, the lower the importance of PDS in selecting suppliers. The null hypothesis could not be rejected; therefore, the hypothesized relationship between the variables was not confirmed nor was any other. 107 Hypothesis 5: The larger the company, either in terms of number of employees or in sales volume, the greater the importance of PDS as a factor in selecting sup- pliers. When company size was measured in terms of number of employees, the null hypothesis was rejected. However, when size was measured in terms of sales volume, the null hypothesis was not rejected, and no other significant relationship between the variables existed. Hypothesis 6: The greater the satisfaction with PDS, the lower its importance as a purchasing factor. The null hypothesis could not be rejected, and no other rela- tionship between the variables was found to be significant. Objective 5: Comparative Analysis To present a comparative analysis of the findings of this study with those reported by Perreault and Russ (1976a). This analysis compares the find- ings in the Brazilian and American environments on the following aspects: (a) satisfaction with over- all PDS and each of its components, (b) importance of PDS as a factor influencing patronage decisions, and (c) relationships involving the perceived impor- tance of PDS in selecting suppliers. The major differences between the two studies centered on the following aspects: --the ranking of satisfaction ratings of the PDS components billing procedures and average delivery time; --the correlation of accuracy in filling orders with satisfac- tion with overall PDS; --the importance ranking of price and PDS as factors in select- ing suppliers; --the relationship between number of deliveries and the impor- tance of P05; 108 --the relationship between satisfaction with overall PDS and its importance in selecting suppliers. The relevant similarities between the two studies were as follows: --the lowest ranking of satisfaction rating with delivery time variability and the higher rankings of both order methods and accuracy in filling orders; --the high correlations of average delivery time and delivery time variability with satisfaction with overall PDS and the fairly low correlations of returns policy and billing procedures with satisfac- tion with overall PDS; --the identical pattern concerning satisfaction with service and feedback about service needs; --the highest importance ranking of product quality and the lowest ranking of reciprocity as factors in selecting suppliers; --the unobserved relationships between number of alternative suppliers and proportion of backorders with importance of PDS in selecting suppliers. The findings of the investigation were presented in this chapter. The interpretation of the findings, implications, and con- clusions are contained in the following chapter. CHAPTER V CONCLUSIONS In this chapter the major conclusions and implications drawn from the findings are presented. The conclusions and implications with respect to satisfaction with overall PDS and its components, the importance of PDS components, and factors affecting the patronage decision are examined in the first three sections. Then, a section on PDS and the patronage decision is presented. The next part deals with the comparative analysis of the findings of this research and those reported by Perreault and Russ (1976a). Finally, the investi- gator proposes further research that might be derived from this study. Satisfaction With PDS The major conclusions and implications concerning satisfac- tion with PDS and its components are presented and discussed below. 1. Respondents who used manufacturers as their major supplier of the chosen product were less satisfied with the overall PDS and with the components average delivery time, rush services, and returns policy than were respondents who used middlemen. For the other compo- nents, the relationships were not significant. Explanations of these findings relate to the underlying characteristics of middlemen and inanufacturers in a channel-of-distribution context: Middlemen tend to be closer to their customers, smaller (at least in the Brazilian 109 110 environment), and more specialized than manufacturers. Moreover, in this research, the average order cycle time of manufacturers (30.6 days) was higher than the average order cycle time of middlemen (9.5 days). Two major implications can be drawn from this conclusion: First, manufacturers could enhance their customers' level of satis- faction, either by improving the overall level of service, or by relying on middlemen to distribute their products, or both. However, two important aspects should be considered in making this decision: The expectations of the customer with respect to the level of service that could be provided either by the middleman or the manufacturer and the cost of providing the service. The outcome might be that manufacturers should rely on middlemen to handle their distribution tasks, especially for highly service-oriented customers. Second, the customers can get better service, all else being equal, either by pressuring manufacturers to provide better services (such as reducing the average order cycle time) or by using middlemen as their major suppliers. 2. The higher the number of alternative suppliers, the lower the satisfaction with the service. Customers seem to require a higher level of performance when the number of suppliers increases. From a supplier perspective, a larger number of competitors would probably necessitate a higher level of service to satisfy the customers' expectations concerning the service provided. 3. The larger companies (measured in terms of number of employees) seemed to be less satisfied with the service provided than were smaller companies. This conclusion should be interpreted with 111 caution, though, because when company size was measured in terms of sales volume, that pattern could not be detected. 4. The higher the percentage of backorders, the lower the satisfaction with overall PDS. Moreover, respondents with no back- orders were significantly more satisfied with the overall PDS and its components (with the exception of billing procedures, for which no pattern was determined) than were respondents with a larger number of backorders. This seems to be an expected finding because number of backorders has a direct effect on the levels of inventory held by the companies. Thus, suppliers can increase the perceived level of satis- faction with PDS by reducing their percentage of backorders. 5. The longer the average order cycle time, the lower the satisfaction with overall PDS. Again, longer order cycle times affect inventory levels; therefore, suppliers can enhance their level of ser- vice by reducing the order cycle time. 6. A considerable number of suppliers sought feedback from their customers with respect to the services they were providing. However, if the revealed expectation was not met, the customers were highly dissatisfied with the service. Therefore, suppliers should be selective in obtaining feedback from their customers if they are not able to meet the customers' expectations about the level of services provided. 7. Even though customers were highly satisfied with the ser- vices they were receiving, a supplier should not neglect the constant monitoring of his customers' satisfaction with the services that he and his competitors are providing. This should enable the supplier 112 to adjust his service offerings to the needs of the customers and to possible competitive actions. 8. Suppliers are constantly faced with the decision of chang- ing PDS levels, either by reducing or increasing their service offer- ings. Since PDS in itself can only be altered by making changes in its components, the issue rests on identifying the effect that each component has on overall satisfaction with PDS, as perceived by customers. Then the cost and revenue implications of altering the service level can be evaluated. This research can provide some insights into this type of decision. Table 5.1 presents the ranking of satisfaction ratings of PDS components and the correlation of each component with the satisfaction with overall PDS. The interpretation of the data in the table is straightforward. A component with a high correlation with satisfaction with overall PDS and a low satisfaction rating is a candidate for improvement (delivery time variability and rush services, for example). An improvement in a component having a low correlation with satisfaction with PDS and a low satisfaction rating (billing procedures and especially returns policy) was deemed to be unsuccessful in increasing the overall satisfaction level of customers with PDS. However, a component with a low correlation and a high satisfaction rating (order methods) could have its level of performance reduced so as to free more funds for increasing perform- ance levels of other components. And components with high correlations and high satisfaction ratings (accuracy in filling orders and average tielivery time) should be closely monitored so that high performance 1«evels can be maintained. All these decisions concerning performance 113 levels should consider the customer's perceptions and consequent reactions to these changes and also the effects on costs and revenues associated with these changes. More meaningful interpretations of the data given in Table 5.1 can be made when the importance rankings of these components are brought into the overall picture. This is done in the following section. Table 5.1.--Groups of satisfaction ratings and correlations. Correlation With Group Component of PDS Mean Overall Satisfaction 1 Order methods 4.134 .37 Actions on complaints 4.124 .56 2 Accuracy in filling orders 4.058 .63 Average delivery time 4.021 .59 3 Order status information 3.978 .53 4 Returns policy 3.896 .34 5 Rush services 3.856 .61 Billing procedures 3.842 .48 Delivery time variability 3.807 .63 All of the interpretations and implications concerning spe- cific actions toward possible improvements in satisfaction with PDS should be considered with caution because the perceived satisfaction of the customers does not necessarily mean any change in patronage decisions (that is, either switching to or from the supplier in ques- tion or remaining with the supplier). 114 importance of PDS Components The mean importance rankings of the PDS components are summar- ized in Table 5.2, in descending order of importance. Table 5.2.--Mean importance rankings of PDS components. .— Order (or group) Component Mean 1 Accuracy in filling orders 2.267 2 Average delivery time 3.126 3 Rush services 3.864 Billing procedures 3.932 4 Actions on complaints 4.576 5 Order status information 4.822 6 Delivery time variability 5.262 7 Returns policy 5.623 Order methods 5.686 The most important PDS component, as perceived by the respond- ents, was accuracy in filling orders. This high ranking was probably a result of the qualifying statement made in the questionnaire about this component: “it means: if the ordered products are delivered with the correct specifications, and at the right time, quantity, and place besides being in usable condition." (See Appendix B for details on the questionnaire.) This statement certainly implies a more encom- passing set of activities associated with the component. Moreover, it is an indication that the respondents perceived the major objective of logistics (". . . to deliver finished inventory and material 115 assortments, in correct quantities, when required, in usable condition, to the location where needed" [Bowersox, 1978, p. 4]) as the most important component of P05. Also, accuracy in filling orders had the highest correlation with satisfaction with PDS (see Table 5.1). This evidence, coupled with the highest ranking of importance of this com- ponent, should constitute a strong argument for possible improvements in the performance level of accuracy in filling orders so as to pro- vide an increase in the satisfaction with PDS and possibly to affect positively patronage decisions. Again, these improvements should be subordinate to forecasted cost and revenue outcomes. Contrary to the theoretical standpoint and to some simulation studies (see Chapter III, section on Research in the Area of PDS), average delivery time was perceived as more important than delivery time variability. This confirms the findings of Ballou (1973)--that buyers apparently cannot discriminate the effects of variability from the average order cycle time. However, respondents from larger com- panies (measured in either sales volume or number of employees) con- sidered delivery time variability more important than average delivery time. An explanation for this finding seems to be twofold: Larger companies have less flexibility in their operations, and they are more inclined to use sophisticated managerial techniques than smaller com— panies. A managerial implication of this finding is that suppliers can augment their service offerings, with respect to delivery time variability, according to the size of their customers. Moreover, delivery time variability also had the highest correlation with satis- faction with PDS (see Table 5.1). However, the importance ranking of 116 this component was very low, which might jeopardize any action toward improving overall satisfaction with PDS. A Since the larger the respondent's company (either in terms of sales volume or number of employees) the higher the importance of delivery time variability, rush services, and accuracy in filling orders; and the lower the importance of billing procedures, suppliers can segment their service offerings according to the size of their customers. Factors Affecting the Patronage Decision Following is a list of conclusions and implications in this area: 1. Product quality was regarded as the most important factor in selecting suppliers. Price was second and PDS third. 2. Respondents who used manufacturers as their major suppliers of the chosen product ranked both product quality and PDS higher in importance among factors in selecting suppliers than did respondents who used middlemen. Explanations of these findings rest in the manu- facturers' possibilities of affecting the quality of their product offerings and in the possibility of suppliers obtaining higher levels of services from manufacturers, when demanded, than from middlemen. However, customers might expect higher performance levels from manu- facturers than from middlemen in terms of both product quality and PDS. 3. With respect to price, the pattern was the opposite; that is, respondents with middlemen as suppliers ranked price higher than did respondents who used manufacturers. It seems that manufacturers 117 are more restrictive in their price policies (usually they are larger companies and not always located close to their customers) than are middlemen. If customers can bargain for lower prices, they certainly consider price as an important factor in their purchasing decisions. 3. Respondents from larger companies regarded PDS as more important in their patronage decisions than did those from smaller companies. Therefore, suppliers should be more concerned with their service offerings when dealing with larger companies. Also, this finding could provide a basis for segmenting service offerings by customer size. PDS and the Patronage Decision The relationships involving the importance of PDS in the patronage decision were formulated in a series of hypotheses presented in preceding chapters. The hypothesized relationships were not found to be significant with reSpect to number of deliveries, proportion of backorders, number of alternative suppliers, or satisfaction with PDS. The hypothesis dealing with order cycle time was confirmed. Thus, the greater the average order cycle time, the greater the impor- tance of PDS in selecting suppliers. According to Perreault and Russ (1976a), "all other things being equal, longer average lead times imply greater variability in lead times, which presumably force the purchaser to evaluate physical distribution service more closely" (p. 6). This explanation was not supported by the findings of this research, since average order cycle time was considered more important than order cycle time variability as PDS components. The hypothesis, however, in itself, 118 could have been derived from the findings dealing with importance of PDS components because average order cycle time was in the second highest group of rankings (see Table 5.2). The implication of the relationship between average order cycle time and importance of P05 is that longer lead times would force the customer to evaluate PDS more closely. With respect to company size and importance of PDS in selecting suppliers, the findings were not conclusive. When company size was measured in terms of number of employees, the relationship was con- firmed. Even though the degree of association was extremely low, the direction of the relationship was meaningful. Thus, the larger the company, the higher the importance of PDS in the patronage decision. However, when company size was measured in terms of sales volume, the relationship was not found to be significant. Despite these two inconclusive findings, smaller companies in both measures of size showed significantly lower importance rankings for PDS as a factor in selecting suppliers. Nevertheless, further implications concerning these findings do not seem appropriate. Even though customers were highly satisfied with PDS, almost all of the hypothesized relationships involving PDS were not signifi- cant, and PDS was ranked third as a factor in selecting suppliers, PDS is still an important element in the interface between the company and its customers. Evidence of this was the high percentage of respond- ents who would switch to another supplier because of a possible stock- out situation. Also, almost half of the respondents had changed sup- pliers in the last two years, and of those, one-quarter had done so 119 because of the poor service they were receiving. Therefore, PDS is indeed an important element of customer service, and both suppliers and customers should be aware that a closer monitoring of services could enhance profitability levels, either by reducing costs or by increasing revenues or both. Comparative Analysis In this section the major aspects of the comparative analysis of the findings of this research and those reported by Perreault and Russ (1976a) are presented and discussed. Satisfaction With PDS and Its Components In essence, the findings of both studies were similar in degree, leading to the conclusion that, with few exceptions, the per- ceptions of satisfaction with PDS and its components can be general- ized for the two environments. Thus, both studies were able to iden- tify the components of PDS whose "improvement is most likely to increase customer satisfaction" (Perreault & Russ, 1976a, p. 10). In the two studies, order cycle time was found to be a candidate for improvement, in both average delivery time and delivery time varia- bility. In contrast, both studies found that improvements in either billing procedures or rush services would not have a significant effect on overall satisfaction with PDS. Importance of Factors in Selecting Suppliers In both studies the subjects were industrial purchasers, and the responses dealt with products that have wide application in 120 manufacturing processes. As expected, respondents in both studies regarded product quality as the most important factor in selecting suppliers. At the other extreme was reciprocity: the least important factor in both studies. The findings with respect to price and PDS were different. In the Brazilian environment, price was more important than PDS in selecting suppliers. In the Perreault and Russ study, PDS "was second only to product quality" (1976a, p. 5). The probable reasons for this difference seem to be twofold: (l) with an economy plagued by high inflation rates, companies tend to hold inventories for specu- latory purposes and as a hedge against inflation; in these circum- stances, price plays an important role in the patronage decision; (2) the concept of physical distribution in Brazil is still in its embrionic stages, and many managers are only beginning to get acquainted with the idea of PDS. Since the importance rankings of the other factors in select- ing suppliers were also different, generalizations about the findings of the two studies should be restricted to product quality and reci- procity. Relationships Similar findings were encountered in the relationship of number of alternative suppliers and percentage of backorders with importance of PDS in selecting suppliers. Both relationships were not significant. With respect to the other relationships, the find— ings of the two studies were different: In the Perreault and Russ 121 study, number of deliveries and satisfaction with overall PDS were significantly related to the importance of PDS in selecting sup: pliers, whereas in this research only average order cycle time was significantly related to importance of PDS in selecting suppliers. Therefore, generalizations can be made only on the similar findings. In conducting this study, the investigator was concerned with the applicability of the findings reported by Perreault and Russ to the Brazilian environment. With the exception of the satisfaction findings, the importance ranking of product quality and reciprocity in selecting suppliers, and the nonexistence of a relationship between number of alternative suppliers and percentage of backorders with importance of PDS in patronage decisions, the other relevant findings cannot be generalized from one environment to the other. Suggested Areas of Further Research Since this research was conducted in the Brazilian environ- ment, the areas of further research were considered in that context and are summarized below: 1. Perceptions of importance of and satisfaction with PDS could be extended to different geographical areas of Brazil, to different industries, and to other product categories. 2. PDS should be examined from a supplier's point of view to determine the level of service offerings and the perceived satisfac- tion with and importance of PDS. Also, the various components of PDS could be identified from the supplier's perspective. 122 3. Since PDS involves tradeoff aspects of cost to service and cost to cost, tradeoff analyses could be used to investigate determinations of adequate levels of service. 4. Additional relationships with PDS as a factor in select- ing suppliers could be examined. For example, the activities per- formed by the executive in charge of purchasing could affect the importance of PDS, and some environmental constraints like inflation and interest rates might influence the importance ranking of PDS and also its different components. 5. Finally, the Perreault and Russ (1976a) framework, with the additions presented in this research, can be replicated in other environments so that additional comparative analyses can be made. APPENDICES 123 APPENDIX A CUSTOMER SERVICE ELEMENTS 124 125 CUSTOMER SERVICE ELEMENTS La Londe and Zinszer (1976, p. 281) - Pretransaction elements “Written statement of policy Customer receives policy statement Organizational structure System flexibility Management services - Transaction elements Stock-out level Order information Elements of order cycle Expedite shipments Transship System accuracy Order convenience Product substitution - Posttransaction elements - Installation, warranty, alterations, repairs, parts - Product tracing - Customer claims, complaints, returns - Temporary replacement of product La Londe and Zinszer (1976, p. 118) - Product availability - Order cycle - Order entry - Order processing - Order picking and shipping - Transit - Distribution flexibility Expedite order Backorder product Substitute product Faster transportation Other 126 - Distribution information Inventory status Order status Data base and forecasting Other - Distribution malfunction - Administrative errors Picking errors Shipping errors Warehouse damage Company-shipping damage Carrier-shipping damage Other - Postsale product support Repair parts availability Repair service Technical advice Other - Other Coyle and Bardy (1980, pp. 346-50) also Gustafson and Richard (1964) - Time Order transmittal Order processing Order preparation Order shipment - Dependability - Lead time - Safe delivery - Correct orders - Communication Convenience Rose (1979, p. 285) - Product availability - Products available when needed — Order completeness - Order accuracy 127 Order cycle - Order processing procedures and time - Order shipments - Order transit times Information services - Inventory reporting - Order status - Data base exchange Order and shipment flexibility - Expedited shipments - Product substitution - Backorder procedures - Alternative transportation service Order and damage adjustments - Adjusting order errors - Correcting shipping errors - Replacing damaged merchandise Product parts and services - Availability of repair parts - Availability of repair service - Technical assistance Bowersox (1978, p. 265) - Capability--order cycle - Availability--inventory levels - Quality Christopher and Walters (1977, p. 56) Availability Delivery Delivery reliability Order processing and progressing Picking errors Back order procedures Returned goods 128 Christopher (1971. PP. 83-87) - Order cycle length - Order transmission - Order preparation - Transport - Consistency of order cycle length - Meeting customer requirements NCPDM--A. T. Kearney (1978. Pp. 188-89) - Order processing - Order entry/editing - Order scheduling - Preparation of order/shipper sets - Invoicing - Customer communications Order modification Order status inquiries Tracing and expediting Error correction Production information requests - Inventory availability and order fill levels - Order cycle times - On-time delivery - Packaging and special handling requirements - Accuracy in all aspects of each customer transaction Perreault and Russ (1976a, p. 8) Billing procedures Average delivery time Delivery time variability Rush service Returns policy Order status information Accuracy in filling orders 129 - Action on complaints - Order methods Perreault and Russ (1974, p. 40) - Order processing time - Order assembly time - Inventory reliability - Order size constraints - Ordering convenience - Delivery time - Consistency - Invoice format - Claims procedure - Inventory backup - Condition of goods - Salesmen's visits - Billing procedures - Order status information - Consolidation allowed Anderson, Jerman, and Constantin (1978, p. 21) — Order cycle time - Transmit the order - Process the order - Ship the material Reliability Damage level Back orders Information systems Stephenson and Willett (1968, p. 78) - Order cycle length Consistency of order cycle length Order preparation Order accuracy Order condition 130 Order size Order frequency Billing accuracy Billing efficiency Back orders Claims Hutchinson and Stolle (1968, p. 88) Order processing time Order assembly time Delivery time Inventory reliability Order-size constraint Consolidation allowed Consistency APPENDIX B QUESTIONNAIRE 131 132 SERVICO PUBLICO FEDERAL UNIVERSIDADE FEDERAL DO RIO GRANDE DO SUL PROGRAMA DE POS-GRADUACAO EM ADMINISTRACAO Prezado Senhor: O Programa de Pos-Graduacao em Administracao da Univer- sidade Federal do Rio Grande do Sul esta iniciando uma pesquisa one term como objetivo principal levantar dados acerca da importancia da prestagao de servigos, pelos fornecedores, como elemento influenciador na obteneao da demanda. Tendo em vista que n50 haveria condicées para serem pesqui- sadas todas as empresas (custo e tempo) foram selecionadas, por amos- tragem, um certo nfimero de empresas para participarem deste estudo. Sua empresa foi uma das escolhidas para colaborar nesta pesquisa. Sua colaboragao, que sera mantida em sigilo, sera efetivada através do me. enchimento do questionario anexo. Na certeza de que vocé compreendera o alcance do trabalho que pretendemos realizar e a importancia destes resultados para sua prépria empresa, a UFRGS espera contar com a sua participacao atra- vés da devolueao do questionario devidamente preenchido até 0 dia ..... /. - . . ./81. Para facilitar seu trabalho encontra-se anexo um en- velope ja selado. Os resultados desta pesquisa serao condensados e analisados em um relatorio que lhe sera remetido Oportunamente. Pela sua compreensao e colaborac‘ao agradecemos antecipada- mente. PROF. FERNANDO BINS LUCE 133 QUESTIONARIO Instrucées para o preenchimento do QUESTIONARIO. E importante que o questiondrio seja respondido pela pessoa a quem foi enderegado, on o responsdvel pelo setor de compras do empresa. Neste questionario vocé encontrara questées referentes a situa- 950 de compra, aos servigos oferecidos por seus fornecedores, a impor- tancia de certos aspectos na escolha de fornecedores, e algumas per- guntas com relacao a sua atitude perante certas situacées. O questionario é de facil preenchimento, basta seguir as ins- trugées contidas no seu interior. Entretanto para facilitar seu trabalho, relacionamos alguns itens, que devem ser observados: 1. Leia com atengao cada questao formulada. 2. Das alternativas fornecidas em cada questao, escolha aquela que melhor represente a sua opiniao em torno do assunto. 3. N50 ha resposta “certa” ou “errada”. 4. A fim de cumprirmos um prazo ja determinado, solicita- mos, dentro do possivel, que o questionario sej a devolvido até 0 dia . . . .. 5. Apes ter respondido a filtima questao, faca uma confe- réncia e verifique se nenhuma deixou de ser respondida. 6. Utilize para devolucao, o envelope que foi enviado em ane- xo. Ele ja esta enderecado e selado basta fechaJo e coloca-lo no correio. 7. O produto a ser escolhido . Abaixo estao selecionados 6 tipos de produtos, genericamente utilizados em processos produtivos . Vocé deve escolher aquele que for de maior importancia (em volume de compras) para sua empresa. Os tipos de produtos foram colocados em ordem, isto é, se vo- cé n50 utilizar na sua empresa o primeiro produto que consta na lista, passe imediatamente para o segundo, e assim por diante, até encontrar aquele que lhe for mais significante. (1) Elementos de fixacao (4) Abrasivos (2) Rolamentos . (5) Eletrodos (3) Lubrificantes (6) Acidos A pesquisa refere-se aos fornecedores, e nao ao produto em si, sendo que estes foram escolhidos exclusivamente como instrumentos do "design” da pesquisa. Assim sendo é importante que vocé nao es- queca que suas respostas dever‘ao ser a respeito do fornecedor mais im- portante em volume de compras do produto escolhido. 134 1. ESCREVA NAS LINHAS PONTILHADAS ABAIXO, QUAL O PRODUTO ESCOLHIDO: ................... ...................................................... 2. SEU FORNECEDOR MAIS IMPORTANTE DESTE PRO- DUTO E: ) Fabricante ) Revendedor ) Distribuidor ) Outro (especifique) ................................. AAAA 3. O NUMERO DE FORNECEDORES UTILIZADOS ...... 4. O NOMERO DE OUTROS FORNECEDORES CONHECI. DOS (nao incluindo os utilizados) ..................... DADOS DE IDENTIFICACAO DA EMPRESA O NI’JMERO DE EMPREGADOS: l a 49 50 a 99 100 a 499 500 a 999 1000 a 4999 5000 on mais 5. ( ( ( ( ( ( vvvvvv VOLUME DE VENDAS NO ANO DE 1980 (em milhfies) de 20 a 49 de 50 a 99 de 100 a 499 de 500 a 999 acima de 1000 vvvvv 7. RAMO PRINCIPAL DE ATIVIDADE ................. 135 8. VOCE EXERCE SUAS FUNQOES EXCLUSIVAMENTE NO SETOR DE COMPRAS? SIM ........ NAO ........ 9- CASO SUA RESPOSTA TENHA SIDO “NAO”, INDI- QUE AS OUTRAS ATIVIDADES QUE VOCE POSSUI DENTRO DA EMPRESA: ............................ I -- SITUAQAO DE COMPRA INSTRUCOES: Escreva nas linhas pontilhadas. “ As questées referem-se ao fornecedor do produto escolhido. “ Os dados fornecidos devem ser do filtimo ano (1980). PERGUNTAS 10. O NI’JMERO DE ENTREGAS, EM-1980: .............. 11. PERCENTAGEM DE PEDIDOS NAO ATENDIDOS NOS PRAZOS ACORDADOS (ou seja pedidos que ficaram pendentes on em carteira). ....... . ................... OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 12. PERCENTAGEM DE PEDIDOS PENDENTES CANCE- LADOS PELA SUA EMPRESA. ..................... 13. DURAQAO MEDIA D0 CICLO D0 PEDIDO (EM DIAS). (intervalo entre a extraefio ou confirmaeao do pedido pelo fornecedor e entrega do produto) ................ II — AVALIAQAO DOS SERVIQOS DE SUPRI- MENTO DO FORNECEDOR Indicagao de seu nivel de satisfacao com relagao aos ser— vicos oferecidos por seu fornecedor mais importante do produto escolhido. 136 INSTRUCAO: Marque um X no interior do parénteses que melhor de- fina o seu grau de satisfagao, com relagao a cada per- gunta. TOTALMEN'I'E TOTALMEN‘I'E ms ”15mm INSATISFEI‘I'O INDIFERENTE smsrerro SATISFEITO l4. Procodimentoo do fatummento: 15. Prazo médio de ontrega: 16. Variabilidado do prazo do entrega: (reiere-se do varia- qées em torno dos prozos médios do ontrega). l7. Pedidos urgentes: 18. Politico do devolu- gées: 19. Informacc'ies sobre o andamento do pedido: 20. Precisao no aiendi- memo do pedido: (precisdo nesto ca- oo signified: so as produto: soliciiados chegam no especifi- cacao, prazo, quan- tidado 9 local com- binados, o em con- dicées do uso). 21. Providéncias toma- daa polo fornecedor om canon do recla- macéo: 23. Métodos do extra- cc‘io do pedidos: 23. Services do supri- monto do fornece- dor: (irate-so do uma avoliocéo a- gregada do todas as Item anterior”). 137 24- ABAIXO ENCONTRAM-SE OS 9 ITENS QUE COM- POE OS SERVIQOS DE SUPRIMENTO DO FORNECE- DOR. ESCOLHA OS 5 ITENS MAIS IMPORTANTES NA SUA OPINIAO, E OS ENUMERE EM ORDEM DE IMPORTANCIA. INSTRUQAO: Coloque os mimeros de 1 a 5 em ordem de importfin- cia, no interior dos parénteses (observe que o nfimero 1 deve ser 0 de maior importéncia). Procedimentos de faturamento: Prazo médio de entrega: Variabilidade no prazo de entrega: Servigos de urgéncia: Politica de devolugoes: Informagoes sobre o andamento do pedido: Precisfio no atendimento do pedido: Providéncias tomadas pelo fornecedor em ca. 803 de reclamagiio: Métodos de extragio de pedidos: AAAAAAAA vvvvvvvv A V III — FATORES INFLUENCIADORES NO PRO- CESSO DE SELECAO DE FORNECEDORES PARA O PRODUTO ESCOLHIDO. 25. INDICACAO DA IMPORTANCIA DE CADA UM DOS FATORES ABAIXO COMO INFLUENCIADORES NA ESCOLHA DE FORNECEDORES. INSTRUCOES: Coloque os nfimeros de 1 a 7 em ordem de importin- cia no interior dos parentéses, (observe que o m'nnero 1 deve ser 0 de maior importfincia). Qualidade do Produto: Servigos de Suprimento do fornecedor: Prego: Geréncia do fornecedor: Localizagfio geogréfica do fornecedor: Pedido minimo requerido: Reprocidade : AAAAAAA vvvvvvv 26. 27. 28- 29. 138 IV — ATITUDES PERANTE CERTAS SITUACOES DE COMPRA; APOS TER SIDO FEITO O PEDIDO DO PRODUTO SE- LECIONADO, SEU FORNECEDOR NOTIFICA QUE O MESMO NAO ESTA DISPONIVEL NO MOMENTO. QUAL A SUA ATITUDE? ........................... OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO ....................................................... ....................................................... ....................................................... DURANTE OS 0LTIMOS 2 ANOS, VOCE TEM MUDA- DO DE FORNECEDORES PARA ESTE PRODUTO? SIM ........ NAO ........ CASO SUA RESPOSTA TENHA SIDO “SIM”, DES- CREVA O MOTIVO DA MUDANCA: ................. ooooooooooooooooooooooooooooooooooooooooooooooooooooooo ooooooooooooooooooooooooooooooooooooooooooooooooooooooo OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OS FORNECEDORES DESTE PRODUTO, CONTAC- TAM COM VOCE OBJETIVANDO SABER SE OS SER- VICOS OFERECIDOS, SAO ADEQUADOS AS SUAS NECESSIDADES ? ( ) Sim, mas melhoramentos 1150 35.0 necessérios. ( ) Sim, e fizeram mudangas que melhoraram as set. vigos. 139 ( ) Sim, mas os servigos nfio melhoraram. ( ) Néo, mas gostaria que fizessem . ( ) Nio, e 1150 é necessério. 30. SE VOCE QUISER FAZER ALGUMA OBSERVAC-AO, UTILIZE O ESPACO ABAIXO. TODAS AS SUAS OPI- NIOES SERAO APRECIADAS. Muito obrigado. 000000000000000000000000000000000000000000000000000000 ...................................................... ...................................................... OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO ...................................................... 140 SERVICO PUBLICO FEDERAL UNIVERSIDADE FEDERAL DO RIO GRANDE DO SUL PROGRAMA DE Pés-GRADUAcKo EM ADMINISTRACKO Dear Sir: The Graduate Program in Business of the Federal University of Rio Grande do Sul--UFRGS--is developing a research project with the main purpose of gathering data on the importance of services pro- vided by suppliers as a demand obtaining factor. Considering that it is impossible to investigate all business companies (cost and time), a certain number of firms were selected, by sampling, to participate in this study. Your firm is one of those selected to collaborate in the research. Your confidential contribu- tion will be effective through a response to the following question- naire. The UFRGS hopes to have your participation and the return of this questionnaire by ..... / ..... /8l because it is counting on your understanding of this study's importance. In order to make the response easier, you will find enclosed a pre-stamped envelope. Research findings and conclusions will be reported to you. Thank you for your support and collaboration. Prof. Fernando Bins Luce 141 QUESTIONNAIRE Directions for answering the questionnaire. It is important that this questionnaire be answered by the person to whom it was addressed, or by the person in charge of the company's purchasing. In this questionnaire you will find questions about the purchas- ing situation, the services offered by suppliers, the importance of certain factors in selecting suppliers, and some items concerning your attitude in specific situations. This questionnaire should be easy to answer; you must only follow directions within it. However, in order to facilitate your job, we outline some gen- eral observations: Read each question carefully. Among the alternatives presented in each question, select the one that best represents your opinion on the subject. There is no "right" or "wrong" answer. To help us meet the schedule, we ask that, whenever possible, you return the questionnaire by ......... (date). After you finish the last question, review to see that you have not missed any one. Use the return envelope enclosed. It is already addressed and stamped; you only have to seal and mail it. 7. The chosen product: Following are six types of products, generally utilized in production processes. You must choose the most important one for your company, in terms of purchasing volume. The types of products listed below are ranked; that is, if you do not utilize the first one on the list, proceed on to the second and so on until you find the most important to you. 0“ 0'1 #00 N-fl (l) fasteners (4) abrasives (2) bearings (5) electrodes (3) lubricants (6) acids The research is on suppliers and not on the products per se, which were chosen solely as a means to achieve the research objec- tives. Thus, it is crucial that you do not forget that your answers refer to the most important supplier, in purchasing volume, of your selected product. 142 2. Your most important supplier of this product is a: ( ) manufacturer ( ) wholesaler ( ) distributor ( ) other (specify): ...................... 3. The number of suppliers utilized is: .............. 4. The number of known alternative suppliers is (do not include the utilized suppliers): ...................... INFORMATION ON YOUR COMPANY 5. Number of employees: ( ) l to 49 ( ) 50 to 99 ( ) 100 to 499 ( 1 500 to 999 ( 1000 to 4999 ( ) 5000 or more 6. Sales volume in 1980 (millions of cruzeiros): ( ) 20 to 49 ( ) 50 to 99 ( ) 100 to 499 ( ) 500 to 999 ( ) over 1000 7. Major business activity .................... 10. 11. 12. 13. 143 Do you work exclusively in the purchasing department? Yes ..... No ..... If your answer was no, state which other responsibilities you have in the company: I. THE PURCHASING SITUATION DIRECTIONS: Write in the blanks. *The questions refer only to the supplier for the chosen product. *Data used must be for the last year (1980). QUESTIONS: The number of deliveries in l980: ............... Percentage of orders. backordered (that is, orders that could not be filled within the requested time): ........... Percentage of backorders canceled by your company: ...... Average order cycle time--in days (that is, the time span between order placement and merchandise delivery) ............ II. EVALUATION OF SUPPLIER'S SERVICES Indicate your satisfaction level concerning the services provided by the most important supplier of the chosen product. 144 Directions: Place an X in the box that best defines your satisfaction level concerning each item. 'O 13 Q1 0) 4—1 .P .P c 9- “- C) U D .F .P .P .P r-H +3 '4- ‘4- F“- r— (U 1‘5 1+- m '- m (U m m 'l— 01- (“U"- H m m 'U +4 44+) Ow- 0!- : (U 0 f5 HO O H U) I—m l4. Billing procedures l5. Average delivery time 16. Delivery time variability (refers to the variations in average delivery time) l7. Rush service l8. Returns policy l9. Order status information 20. Accuracy in filling orders (it means: if the ordered products are delivered with the correct specifications and at the right time, quantity, and place, besides being in usable con- dition) 21. Actions taken by supplier in case of complaints 22. Order methods 23. Suppliers distribution services (that is, an aggregated evaluat1on of all items above) 145 24. Following there are the nine items that constitute the suppliers' physical distribution services. Among them, choose the five you find most important and rank them in order of importance. DIRECTIONS: Choose the five most important items and rank them from l to 5 in order of importance (note that l is the most important). billing procedures average delivery time delivery time variability rush service returns policy order status information accuracy in filling orders actions taken by supplier in case of complaints order methods ( ( ( ( ( ( ( ( ( III. FACTORS INFLUENCING THE PROCESS OF SUPPLIER SELECTION FOR THE CHOSEN PRODUCT 25. For each of the factors listed below, indicate the importance they have in influencing the selection of suppliers. DIRECTIONS: Rank the following factors from 1 to 7 in order of importance (note that l is the most important). ( ) product quality ( ) supplier's distribution services ( ) price ( ) supplier's management ( ) supplier's location ( ) minimum required order size ( ) reciprocity 26. 27. 28. 29. 146 IV. ATTITUDES ON CERTAIN PURCHASING SITUATIONS After you place an order for the chosen product, your supplier notifies you that it is not available at the moment. What is your attitude? During the last two years, have you changed suppliers for the chosen product? Yes ...... No ....... If your answer was yes, describe the reasons for the change: Do your suppliers of the chosen product check with you to see if the services they are providing are adequate in meeting your needs? ( ) Yes, but there are no needed improvements. ( ) Yes, and they have made improvements. ( ) Yes, but the services did not improve. ( ) No, but I would like them to do so. ( ) No, and it is not necessary. 147 30. If you want to make any additional observations, please use the blanks below. All your comments will be appreciated. Thank you. APPENDIX C LIST OF VARIABLES 148 149 LIST OF VARIABLES SITUATIONAL VARIABLES Vl5--Number of deliveries (Question l0) Vl6--Percentage of backorders (Question ll) Vl7--Percentage of backorders cancelled (Question 12) V18--Average order cycle time (Question 13) SUPPLIER VARIABLES V2--Category of supplier (Question 2) V3—-Number of suppliers utilized (Question 3) V4--Number of other suppliers (Question 4) V50--Number of alternative suppliers (V3 + V4) COMPANY VARIABLES V5--Size by number of employees (Question 5) V6--Size by sales volume (Question 6) V7--Type of industry (Question 7) SATISFACTION VARIABLES Vl9--Billing procedures (Question 14) V20--Average delivery time (Question l5) V21--Delivery time variability (Question l6) V22--Rush services (Question 17) V23--Returns policy (Question l8) V24—-Order status information (Question l9) V25--Accuracy in filling orders (Question 20) 150 V26--Actions on complaints (Question 21) V27--Order methods (Question 22) V28--Overall PDS (Question 23) V5l--Overall PDS (weighted average of satisfaction of com- ponents by their respective importance) IMPORTANCE VARIABLES (Question 24) V29--Billing procedures V30--Average delivery time V3l--Delivery time variability V32--Rush services V33--Returns policy V34--Order status information V35--Accuracy in filling orders V36--Actions on complaints V37--Order methods PURCHASING FACTORS VARIABLES (Question 25) V38--Product quality V39--Physical distribution service (PDS) V40--Price V4l--Supplier management V42—-Geographical location of supplier V43—-Required minimum order size V44--Reciprocity 151 RESPONDENT VARIABLES V8--Purchasing function (Question 8) V9 to Vl4--Other functions or activities (Question 9): V9--General management VlO--Finance Vll--Production Vl2--Personnel Vl3--Marketing Vl4--Materials management PRODUCT VARIABLE Vl--The chosen product (Question 1) GENERAL VARIABLES V45--Attitudes toward possible stockout (Question 26) V46--Changes in suppliers (Question 27) V52 to V55--Reasons for changing suppliers (Question 28): V52--Price V53--Product quality V54--PDS V55--0ther V48--Feedback on services (Question 29) V49--Comments (Question 30) BIBLIOGRAPHY 152 BIBLIOGRAPHY Anderson, Ronald 0.; Jerman, Roger E.; and Constantin, James A. "Structure and Analysis of Physical Distribution Goals." Journal of Business Logistics 1,1 (1978): 19-30. Anderson, Ronald 0.; Jerman, Roger E.; and Maister, David H. "Alter- native Approaches to the Evaluation of Physical Distribution Activities: Preferences of Executives in Canadian Firms." Transportation Journal 18 (Summer 1979): 64-71. Ballou, Ronald H. Business Logistics Management. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1973. . “Planning a Sales Strategy With Distribution Service." lhe Logistics and Transportation Review 9,4 (1973): 323-34. . Basic Business Lo istics. Englewood Cliffs, N.J.: Prentice Hall, Inc., 1 8. "An Extended Distribution Analysis to Support Marketing and Production Planning." Journal of Business Logistics 1,1 (1978): 63-75. , and De Hayes, Daniel W., Jr. "Transport Selection by Inter- firm Analysis." Transportation and Distribution Management 7 (June 1967): 33-37. Banting, Peter M. "Customer Service in Industrial Marketing: A Comparative Study." European Journal of Marketing 10,3 (1976): 136-45. Bartlett, Hale C. Readings in Physical Distribution. Danville, 111.: The Interstate Printers & Publishers, Inc., 1970. Bender, Paul 5. Design and Operation of Customer Service Systems. New York: AMACOM, 1976. Bleuel, William H., and Bender, Henry E. Product Service Planning. New York: AMACOM, 1980. Bowersox, Donald J. Logistical Management. New York: Macmillan Publishing Co., Inc., 1978. 153 154 Brunner, G. Allen, and Carroll, Stephen J., Jr. “The Effect of Prior Notification on the Refusal Rate on Fixed Address Surveys." Journal of Advertising Research 9 (March 1969): 42-44. Cavinato, Joseph L., and Perreault, William 0., Jr. “Evaluating Vendors' Distribution Service.“ Journal of Purchasing_and Materials Management 12 (Summer 1976): 13-19. Christopher, Martin. Total Distribution. Westmead, Farnborough, Hampshire, England: Gower Press, Teakfield, Ltd., 1971. , and Gattorna, John. Controlling the Distribution Function. Bradford, West Yorkshire, England: MCB Publications, n.d. Christopher, Martin; Schary, Philip; and Skjott-Larsen, Tage. Customer Service and Distribution Strategy. New York: John Wiley & Sons, 1979. Christopher, Martin, and Walters, David. Distribution Planning and Control. Westmead, Farnborough, Hants, England: GBWer Press, Teakfield, Ltd., 1977. Christopher, Martin, and Wills, Gordon. "Developing Customer Service Policies." Physical Distribution Monograph 4,6 (1974): 321-52. Coyle, John H., and Bardi, Edward J. The Management of Business Logistics. St. Paul, Minn.: West Publishing Co., 1980. Cunningham, M. T., and Hardy, S. M. R. "Evaluating the Effectiveness of Sales and Distribution Systems." In Controlling the Distri- bution Function, pp. 106-33. Edited by Martin Christopher and John Gattorna. Bradford, England: MCB Publications, n.d. , and Roberts, 0. A. "The Role of Customer Service in Market- ing." European Journal of Marketing_8 (Spring 1974): 15-28. Cunningham, M. T., and White, J. G. "The Determinants of Choice of Supplier." European Journal of Marketing 7 (Winter 1973/74): 189-202. Daniel, Norman E., and Jones, J. Richard. Business Logistics. Boston: Allyn & Bacon, Inc., 1969. Davis, Herbert W. "Four Reasons Why Customer Service Managers Can't Mana e Customer Service." Handling and Shipping 12 November 1971): 51-53. Ferriss, Abbot L. "A Note on Stimulating Response to Questionnaires." American Sociological Review 16 (April 1951): 247-49. 155 Finney, F. Robert. "Reciprocity: Gone But Not Forgotten." Journal of Marketing_42 (January 1978): 54-59. Gilmour, Peter. "Development of a Demand Response Function." Journal of Business Logistics 1,2 (1979): 83-102. Guilford, J. P. Fundamental Statistics in Psychology and Education. New York: McGraw-Hill'BOOk Co.,Pl965. Gustafson, John F., and Richard, Raymond. "Customer Service in Physical Distribution. In Business Logistics, pp. 349-56. Edited by Norman E. Daniel and J} Richard Jones. Boston: Allyn & Bacon, Inc., 1969. Herron, David P. "Managing Physical Distribution for Profit." Harvard Business Review 57 (May-June 1979): 121-32. Heskett, James L. "Predictive Value of Classroom Simulation." In Emerging Concepts in Marketing, pp. 101-15. Edited by William S. Decker. Chicago: American Marketing Association, 1963. . "Controlling Customer Logistics Service." International Journal of Physical Distribution (June 1971): 141-45. ; Glaskowsky, Nicholas A., Jr.; and Ivie, Robert M. Business Eogistics. New York: The Ronald Press Co., 1973. Hutchinson, William J., Jr., and Stolle, John F. "How to Manage Customer Service." Harvard Business Review 46 (November- December 1968): 85-96. Johnson, Richard A., and Parker, Donald 0. "Optimizing Customer Delivery Service With Improved Distribution." Business Review 21 (October 1961): 38-46. Kerlinger, Fred N. Foundations of Behavioral Research. New York: Holt, Rinehart, & Winston, Inc., 1973. Kirk, Roger E. Experimental Design: Procedures for the Behavioral Sciences. Belmont, Ca1if.: Wadsworth PDblishing Co., Inc., 1968. Klass, Bertrand. "What Factors Affect Industrial Buying Decisions." Industrial Marketing 47 (May 1961): 33-35. Kotler, Philip. Marketing Management. Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1967. La Londe, Bernard J., and Zinszer, Paul H. Customer Service: Meaning and Measurement. Chicago: National Council of Physical Distri- bution Management, 1976. 156 Lehmann, Donald R., and O'Shaughnessy, John. "Difference in Attri- bute Importance for Different Industrial Products." Journal of Marketing 38 (April 1974): 36-42. Levy, Michael. "Toward an Optimal Customer Service Package." Journal of Business Logistics 2,2 (1981): 87-109. Luce, Fernando 8.; Fensterseiffer, Jaime; and Hoppen, Norberto. “Custos de Manutencao de Estoques." Documentos de Estudo PPGA/ UFRGS, Porto Alegre/RS, forthcoming 1982. Magee, John F. "The Logistics of Distribution." Harvard Business Review 38 (July-August 1960): 89-101. . Physical-Distribution Systems. New York: McGraw-Hill Book Co., 1967. Moseley, Lloyd W. Customer Service. New York: Chain Store Publish- ing Corp., 1972. Myers, James H., and Haug, Arne F. "How a Preliminary Letter Affects Mail Survey Returns and Costs." Journal of Advertising Research 9 (September 1969): 37-39. NCPDM. Measuring Productivity in Physical Distribution. Chicago: National Council of Physical Distribution Management, 1978. Nie, Norman H.; Hull, C. Hadlai; Jenkins, Jean13.;Steinbrenner, Karin; and Bent, Dale H. Statistical Packa e for the Social Sciences. New York: McGraw-Hill Book Co., 197 . Perreault, William 0., Jr. "The Role of Physical Distribution Cus- tomer Services in Industrial Purchase Decisions." Ph.D. disser- tation, University of North Carolina, 1973. , and Russ, Frederick A. "Physical Distribution Service: A Neglected Aspect of Marketing Management." MSU Business Topics (Summer 1974): 37-45. . "Physical Distribution Service in Industrial Purchase Decisions." Journal of Marketing 40 (April 1976): 3-10. . "Quantifying Marketing Trade-offs in Physical Distribution Policy Decisions." Decision Sciences 7 (April 1976): 186-201. . “Improving Physical Distribution Service Decisions With lrade-off Analysis." International Journal of Physical Distri- bution 7,3 (1976): 117-27. 157 Ray, David, and Millman, Stephen. "Optimal Inventories via Customer Service Objectives." International Journal of Physical Distri- bution and Materials Management 9,7 (1979): 325-49. Rose, Warren. Logistics Management. Dubuque, Iowa: Wm. C. Brown Co., Publishers, 1979. Schary, Philip B., and Becker, Boris W. "Product Availability and the Management of Demand." European Journal of Marketing 10,3 (1976): 127-35. Scott, C. "Research on Mail Questionnaires." Journal of the Royal Statistical Society 124, series A (1961): 143-92. Semon, Thomas T. "A Cautionary Note on 'Difference in Attribute Importance for Different Industrial Products.'" Journal of Marketing 39 (January 1975): 79. Shycon, Harvey N., and Sprague, Christopher R. “Put a Price Tag on Your Customer Servicing Levels." Harvard Business Review 53 (July-August 1975): 71-78. Siegel, Sidney. Nonparametric Statistics for the Behavioral Sciences. New York: McGraw-Hill Book Co., Inc.,Pl956. Simon, Leonard 5. "Measuring the Market Impact of Technical Services." Journal of Marketing Research 2 (February 1965): 32-39. Spem, Thomas, and Wagenheim, George 0. "Demand and Lead-Time Uncer- tainty: The Impacts on Physical Distribution Performance and Management." Journal of Business Logistics 1,1 (1978): 95-113. Stephenson, Paul R. "Manufacturers' Physical Distribution Service Knowledge and Penalties: An Experimental Analysis. In Pa ers-- 4th Annual Meeting, Transportation Research Forum (1963), pp. 62-71. , and Willett, Ronald P. “Selling With Physical Distribution Services." Business Horizons 11 (December 1968): 75-85. Stolle, John F. "How to Manage Customer Service." Harvard Business Review 45 (July-August 1967): 93-100. Uhr, Ernest B.; Houck, Ernest C.; and Rogers, John C. "Physical Distribution Service." Journal of Business Logistics 2,2 (1981): 158-69. van Buijtenen, Pieter. Business Logistics. The Hague, the Nether- lands: Martinus NijhOff, 1976. 158 Vocino, Thomas. "Three Variables in Stimulatin Responses to Mailed Questionnaires." Journal of Marketing_41 ?0ctober 1977): 76-77. Voorhees, Roy Dale, and Sharp, Merrill Kim. "The Principles of Logis- tics Revisited." Transportation Journal 18 (Fall 1978): 69-84. Waisanen, F. B. "A Note on the Response to a Mailed Questionnaire.” Public Opinion Quarterly 18 (Summer 1954): 210-12. Webster, Frederick E., Jr. "Modeling the Industrial Buying Process." Journal of Marketing Research 2 (November 1965): 370-76. Weigand, Robert E. "Identifying Industrial Buying Responsibilities." Journal of Marketing Research 3 (February 1966): 81-84. Wentworth, Felix. Handbook of Physical Distribution Management. Epping, Essex, England: *Gower Press, Ltd., 1976. White, Phillip 0., and Cundiff, Edward W. "Assessing the Quality of Industrial Products." Journal of Marketing 42 (January 1978): 80-86. Willett, Ronald P., and Stephenson, P. Ronald. "Determinants of Buyer Response to Physical Distribution Service." Journal of Marketing Research 6 (August 1969): 279-83. MICHIGAN sms UNIV. LIBRARIES 1111111111111111111111111111111111 31293104991181