.IIIIIIIIIIIIIIli IIIIIII IIIIIIII , IIIIIIIIII IIIII IIIIIIII IIIIII ”BRA“? 293 10416 0860 Michigan §tate University *- IL This is to certify that the dissertation entitled TOWARD AN IMPROVED DISTRIBUTION ACCOUNTING INFORMATION SYSTEM THROUGH AN ENTITY—RELATIONSHIP DATA MODELING APPROACH presented by Howard MacKay Armitage has been accepted towards fulfillment of the requirements for Ph.D. Accounting degree in WU“ I Major professor Date gfa 63 MS U it an Affirmative Action/Equal Opportunity Institution 0 .12771 MSU LIBRARIES 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. TOWARD AN IMPROVED DISTRIBUTION ACCOUNTING INFORMATION SYSTEM THROUGH AN ENTITY-RELATIONSHIP DATA MODELING APPROACH By Howard MacKay Armitage A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Accounting 1983 A B S T R A C T TOWARD AN IMPROVED DISTRIBUTION ACCOUNTING INFORMATION SYSTEM THROUGH AN ENTITY-RELATIONSHIP DATA MODELING APPROACH by Howard MacKay Armitage The integrated physical. distributitnl management concept has received widespread support in the business and academic communities. Evidence suggests, however, that operationalizing the concept has been hindered by a lack of adequate cost data. Data provided by the accounting system is generally too highly aggregated to be of use in distribution management and ad hoc data gathering efforts to overcome these limitations lead to cost inefficiencies and the absence of a shared data resource. Attempts by first generation data models to improve information availability by' overlaying the existing accounting structure with data base technology have proven unsatisfactory. Based on the events theory of accounting, this dissertation utilizes a structured analysis approach to obtaining user require- ments and a data model, the Entity-Relationship model, to develop an integrated conceptual information schema in a distribution setting. The methodology' is applied to an research situation consisting of an existing system and three separate cases of new distribution requirements which the literature indicates have been previously unsupported by the traditional accounting model. The results of the study indicate that the methodology developed in the research can be used to successfully identify, model and integrate the requirements of a large distribution accounting system. The conceptual schema obtained in the study represents a consolidation of views which meets both the financial needs of accountants and the managerial needs of several levels of distribution executives. Implications of the methodology for both accountants and distribution personnel are presented anui future research directions are outlined. This dissertation is dedicated to those three individuals who shared all the joys and traumas of the doctoral program with me. For their unending source of love, strength and inspiration, this dissertation is dedicated to my wife Phyllis and my children, Alanna and Derek. ii ACKNOWLEDGEMENTS Without the support of many different individuals and groups, this dissertation would never have been completed. First, I would like to express my sincere gratitude and appreciation to my committee ‘members. Bun Harold Sollenberger provided sound guidance and council throughout my program. His advice, particularly during the early stages of the dissertation, was often sought and always greatly appreciated. ‘Dr. Douglas Lambert, my co-chairman, first introduced me to the richness of the distribution accounting area. It was his knowledge and enthusiasm that stimulated my interest in this field. His many helpful and insightful comments during the writing of the dissertation were most appreciated as were the numerous joint academic and professional opportunities that he initiated. Dr. Willianl McCarthy, my other co-chairman provided expert council throughout the entire project. His concern, involvement, intellect and integrity were a driving force behind the research. For this immeasurable contribution any acknowledgement seems too meagre. Perhaps it is enough to say that his direction in the research was always helpful, always useful and always highly valued. iii Second, I would like to acknowledge the generous financial support of the National Council. of Physical Distribution. whose desire to improve distribution accounting procedures was a major motivation for pursuing the research. Special thanks go to Ursula Kirch and Ann Puncher who courageously, and at times overcoming all odds, bore the responsibility for typing the manuscript. Finally, I would like to acknowledge the support of Dr. Jack Hanna and the Centre for Accounting Research at the University of Waterloo for providing the time and financial support to permit the completion of the research. iv TABLE OF CONTENTS CHAPTER I INTRODUCTION Background The Business Problem The Research Purpose The Research Objectives Scope and Limitations Value of the Study Organization References CHAPTER II REVIEW OF THE LITERATURE Review of the Distribution Accounting Literature The Role of Accounting in Distribution Management The Generalized Accounting Problem Review of the Events Accounting Literature and Its Early Applications Theory Development Implementing the Events Approach - The First Applications Hierarchical Events Systems Relational Events Systems Summary of the First Applications Review of the Entity-Relationship Data Model and an Accounting Data Model Extension Review of Database Design Literature Relevant to the Research Requirements Analysis Information Analysis and Definition View Modeling and Modification View Analysis and Integration Summary References 17 18 25 28 30 33 33 39 4O 41 48 49 SS 56 56 57 S9 CHAPTER III RESEARCH DESIGN Methodological Overview Methodology Used to Obtain and Analyze Information Requirements in This Study New Requirements Analysis Existing Requirements Analysis Advantages Associated with Research Design The Research Situation New Requirements Analysis Accounting for Warehouse Labor Productivity Statement of Requirements Analysis of Requirements Choosing the Least Cost Distribution Center Statement of Requirements Analysis of Requirements The Warehouse Location Decision Statement of Requirements Analysis of Requirements Summary of New Requirements Existing Requirements Anslysis A Physical View of the Existing Distribution Information System A Logical View of the Existing Distribution Information System Conclusion Appendix A Appendix B Appendix C References CHAPTER IV MODELING THE DISTRIBUTION ACCOUNTING SYSTEM ‘View Modeling and Modification View Modeling and Modification - Case I Data Modeled Views Data Element Dictionary Transform Descriptons The New Logical Data Flow Diagram Summary of Case 1 View Modeling and Modification - Case II Data Modeled Views Data Element Dictionary, Transform Descriptions and New Logical Data Flow Diagram vi Page 65 69 7O 74 75 76 81 82 84 86 94 97 98 106 107 110 113 115 116 122 128 129 145 154 171 174 177 178 202 213 218 221 221 221 236 View Modeling and Modification - Case III Data Modeled Views View Modeling and Modification - The Existing System Data Modeled Views View Analysis and Integration View Analysis and Integration - Declarative Phase Schema Specification The New Logical Data Flow Diagrams View Analysis and Integration - Procedural Phase Transform Descriptions Conclusion References CHAPTER V SUMMARY Summary of Research Motivation and Research Objectives Methodologies and Research Design Used in the Study Review of the Results and Business Implications Accounting Implications Distribution Implications Issues Not Addressed by the Dissertation Conclusion and Directions for Future Research References BIBLIOGRAPHY List of References vii Page 236 242 252 253 263 264 264 275 287 287 291 293 294 295 297 299 301 303 304 306 307 10. 11. 12. 13- LIST OF TABLES ESTIMATES OF PHYSICAL DISTRIBUTION COSTS EVENTS ACCOUNTING ARTICLES TABLE DESCRIBING VARIOUS PROFILES OF THE COMPANY'S BUSINESS OPERATIONS INTERVIEWS CONDUCTED TO DETERMINE THE CRITICAL FACTORS INVOLVED IN LOCATION DECISIONS DESCRIPTION OF SYMBOLS USED IN THE STUDY TABLE ILLUSTRATING ATTRIBUTES OF ENTITY AND RELATIONSHIP SETS FOR CASE 1 "THE WAREHOUSE LABOR AND PRODUCTIVITY SYSTEM" DATA ELEMENT DICTIONARY ENTRIES FOR DATA FLOWS WHICH TAKE PLACE WHEN EMPLOYEES HANDLE FINISHED GOODS DATA ELEMENT DICTIONARY ENTRIES FOR DATA FLOWS WHICH TAKE PLACE WHEN TRANSACTION DETAILS ARE TRANSFORMED INTO PRODUCTIVITY AND COST REPORTS TRANSFORM DESCRIPTIONS OF PROCESS 1.3 to 1.8 TRANSFORMATION OF TRANSACTION DETAIL DATA INTO PRODUCTIVITY AND COST REPORTS - PROCESS 2.6 TABLE ILLUSTRATING ATTRIBUTES OF ENTITY AND RELATIONSHIP SETS FOR CASE 2 "THE LEAST COST DISTRIBUTION DECISION" TABLE ILLUSTRATING ATTRIBUTES OF ENTITY AND RELATIONSHIP SETS FOR CASE 3 "THE SITE SELECTION DECISION" RELATIONAL ALGEBRA EXAMPLE ILLUSTRATING PROCEDURES FOR OBTAINING TASK AND WORK AREA PRODUCTIVITY REPORTS viii PAGE 20 29 79 111 175 204 209 211 215 216 238 250 289 LIST OF TABLES cont'd. TABLE A1. A2. C1. TABLE ILLUSTRATING THE RESULTS OF AN INDUSTRIAL ENGINEERING STUDY TO DETERMINE THE TIME REQUIRED TO COMPLETE VARIOUS TASKS IN THE RECEIVING AND STORING, PICKING, PACKING AND SHIPPING WORK AREAS OF REPRESENTATIVE PFW's, RDC's, and LDC's TABLE ILLUSTRATING RECEIVING AND STORING PRODUCTIVITY EFFICIENCY MEASURES FOR REPRESENTATIVE PRODUCTION FACILITY WAREHOUSE IN THE REGIONAL DISTRIBUTION SYSTEM HIERARCHY OF LOGICAL DATA FLOW DIAGRAMS ix PAGE 137 139 155 9a. 9b. 10. 11. 12. 13. 14. 15. 16. LIST OF FIGURES COST TRADEOFFS REQUIRED IN A PHYSICAL DISTRIBUTION SYSTEM A SALES MANAGER'S CONCEPTUAL VIEW OF A DATABASE AN ENTITY-RELATIONSHIP DIAGRAM ILLUSTRATING NATURE OF RELATIONSHIPS BETWEEN ENTITY SETS ATTRIBUTE-VALUE SETS ON ENTITY SET "VENDOR" "VENDOR" ENTITY-RELATION TABLE DIAGRAM ILLUSTRATING SEQUENCE OF DATABASE DESIGN PROCEDURES COMPONENTS OF STRUCTURED ANALYSIS PROJECT LIFE CYCLE SURVEY DATA FLOW DIAGRAM STRUCTURED ANALYSIS DATA FLOW DIAGRAM DIAGRAM ILLUSTRATING COMPONENTS OF RESEARCH DESIGN DIAGRAM ILLUSTRATING FLOW OF PRODUCT FOR A PHYSICAL DISTRIBUTION NETWORK CONTEXT DIAGRAM OF A WAREHOUSE LABOUR REPORTING SYSTEM FOR RECEIVING AND STORING DIAGRAM 1: RECEIVING AND STORING DIAGRAM 2: PAYROLL AND LABOR DISTRIBUTION DIAGRAM O: EVALUATE DISTRIBUTION ALTERNATIVES DIAGRAM 4: TRAFFIC PAGE 19 37 44 45 46 51 53 66 67 67 72 78 89 9O 91 101 102 LIST OF FIGURES 17. 18. 19. 208. 20b. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. FIGURES cont'd. DIAGRAM 5: INVENTORY CARRYING COSTS DIAGRAM ILLUSTRATING EXISTING AND PROPOSED WAREHOUSE SITE CONFIGURATION DIAGRAM OF THE DISTRIBUTION CENTER SITE SELECTION PROCESS CONTEXT DIAGRAM: TRANSPORTATION COSTS OBTAIN INBOUND AND OUTBOUND TRANSPORTATION COSTS ORDER PROCESSING ORDER RECEIVING AND PROCUREMENT PAYROLL AND LABOR DISTRIBUTION ORDER RECEIVING AND PROCUREMENT RECEIVING AND STORING FILE UPDATE AND PROCUREMENT PAYROLL AND LABOR DISTRIBUTION THE INFORMATION ANALYSIS AND DEFINITION PHASE OF DESIGN DIAGRAM 1: RECEIVING AND STORING DIAGRAM 2: PAYROLL AND LABOR DISTRIBUTION E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.1 "MATCH RECEIPT DOCUMENT TO ACQUISITION ORDER" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.2 "PREPARE WORK ACTIVITIES" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.3 "ASSIGN EMPLOYEES TO TASKS" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.4 "MARK TIME THAT TASK BEGINS" xi PAGE 103 109 112 114 114 117 120 121 124 125 126 127 176 179 180 181 185 187 188 LIST OF FIGURES cont'd. FIGURES 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. E-R DIAGRAM OF INFORMATION REQUIRED FOR PROCESSES 1.5 to 1.8 "COMPLETE TASKS" E-R DIAGRAM OF INFORMATION REQUIRED FOR PROCESS 1.9 "PREPARE RECEIVING REPORT" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.10 "PERFORM CONSISTENCY CHECK" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.11 "RECORD EMPLOYEE DEPARTURE" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.12 "COMPARE EMPLOYEE TIMES WITH DAILY TASK RECORDERS" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 2.1 "VERIFICATION AND EDIT" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 2.2 "UPDATE FILES" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 2.3 "PREPARE PAYROLL REPORTS AND PRINT CHECKS" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 2.4 "PREPARE YEAR-END TAX REPORTS" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 2.5 "PREPARE TRANSACTIONS DETAIL E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 2.6 "PREPARE PERFORMANCE REPORTS" E-R DIAGRAM OF THE ENTITIES AND RELATIONSHIPS ASSOCIATED WITH CASE 1 "WAREHOUSE LABOR PRODUCTIVITY SYSTEM" DIAGRAM 1: RECEIVING AND STORING DIAGRAM 2: PAYROLL AND LABOR DISTRIBUTION COMPARISON OF EXISTING AND NEW SYSTEM REQUIREMENTS "RECEIVING AND STORING" xii PAGE 189 191 192 193 194 194 196 198 199 200 201 203 219 220 222 LIST OF FIGURES 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62a. 62b. 63. FIGURES cont'd. COMPARISON OF NEW AND EXISTING SYSTEM REQUIREMENTS "PAYROLL AND LABOR DISTRIBUTION" DIAGRAM O: EVALUATE DISTRIBUTION ALTERNATIVES DIAGRAM 4: TRAFFIC DIAGRAM 5: INVENTORY CARRYING COSTS E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESSES 1, 2 and 3 "RECEIVING AND STORING, FILL ORDER AND LABOR DISTRIBUTION" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 4.1 "MATCH INVOICE TO BILL OF LADING" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 4.2 "COMPUTE FREIGHT COSTS" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 5.1 "OBTAIN INVENTORY CARRYING COSTS BY PRODUCT BY LOCATION" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 5.2 "DETERMINE AVERAGE INVENTORY CARRYING COSTS BY PRODUCT BY LOCATION" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 5.3 "DETERMINE AVERAGE INVENTORY CARRYING COSTS BY PRODUCT IN TRANSIT" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 6 "OBTAIN CUSTOMER ORDER DETAILS" E-R DIAGRAM OF THE ENTITIES AND RELATIONSHIPS ASSOCIATED WITH CASE 2 "THE LEAST COST DISTRIBUTION SYSTEM" CONTEXT DIAGRAM: TRANSPORTATION COSTS OBTAIN INBOUND AND OUTBOUND TRANSPORTATION COSTS E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.1 "CONSOLIDATE PRIVATE CARRIAGE TRIP COSTS" xiii PAGE 223 224 225 226 228 230 232 233 233 233 235 237 243 243 244 LIST OF FIGURES 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76a. 76b. 77. 78. FIGURES cont'd. E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.2 "ALLOCATE TRIP COST TO SHIPMENTS WITHIN TRIP" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.4 "MATCH INVOICE TO BILL OF LADING" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.5 "ALLOCATE COMMON CARRIER FREIGHT COSTS TO SHIPMENTS" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.7 "MATCH UPS INVOICE" E-R DIAGRAM OF INFORMATION REQUIRED BY PROCESS 1.8 "DETERMINE UPS FREIGHT COSTS BY SHIPMENT" E-R DIAGRAM OF THE ENTITIES AND RELATIONSHIPS ASSOCIATED WITH CASE 3 "THE SITE SELECTION DECISION" E-R DIAGRAM OF THE EXISTING ORDER PROCESSING SYSTEM E-R DIAGRAM OF THE EXISTING ORDER RECEIVING AND PROCUREMENT SYSTEM E-R DIAGRAM OF THE EXISTING TRAFFIC SYSTEM E-R DIAGRAM OF EXISTING ACCOUNTS RECEIVABLE SYSTEM E-R DIAGRAM OF THE EXISTING PAYROLL AND LABOR DISTRIBUTION SYSTEM E-R DIAGRAM OF EXISTING ACCOUNTS PAYABLE SYSTEM REA GENERAL MODEL SPECIFIC APPLICATION OF THE REA MODEL FINAL CONCEPTUAL SCHEMA INTEGRATING NEW REQUIREMENTS WITH EXISTING SYSTEM FIRST INTEGRATION OF LOCAL EMPLOYEE VIEWS xiv PAGE 245 246 247 248 248 249 254 256 257 259 260 262 266 266 270 271 LIST OF FIGURES 79. 80. 81. 82. 83. 84. 85. 86. 87. A1. B1. B2. B3. B4. B5. B6. B7. C1. C2. C3. C4. FIGURES cont'd. SECOND INTEGRATION OF LOCAL EMPLOYEE VIEWS DIAGRAM O: DISTRIBUTION ACCOUNTING DIAGRAM 1: DISTRIBUTION INFORMATION PROCESSING DIAGRAM 1.2: ORDER RECEIVING AND PROCUREMENT DIAGRAM 1.2.1: RECEIVING AND STORING DIAGRAM 2: ACCOUNTING INFORMATION SYSTEM DIAGRAM 2.2: PAYROLL AND THE LABOR DISTRIBUTION DIAGRAM 1.2.1: RECEIVING AND STORING DIAGRAM 2.2: PAYROLL AND LABOR DISTRIBUTION DIAGRAM ILLUSTRATING SEQUENCE OF PROCEDURES AT INVENTORY LOCATIONS ORDER PROCESSING ORDER RECEIVING AND PROCUREMENT TRAFFIC ACCOUNTS RECEIVABLE PAYROLL AND LABOR DISTRIBUTION ACCOUNTS PAYABLE GENERAL LEDGER CONTEXT DIAGRAM DIAGRAM O: DISTRIBUTION ACCOUNTING DIAGRAM 1: DISTRIBUTION INFORMATION PROCESSING DIAGRAM 1.1: ORDER PROCESSING PAGE 273 279 280 281 282 283 284 285 286 133 147 148 149 150 151 152 153 156 157 158 159 C5. C6. C7. C8. C9. C10. C11. C12. C13. C14 0 C15. DIAGRAM 1.1.1: ENTER ORDER DIAGRAM 1.1.2: FILL ORDER DIAGRAM 1.2: ORDER RECEIVING AND PROCUREMENT DIAGRAM 1.2.1: RECEIVING AND STORING DIAGRAM 1.2.2: FILE UPDATE AND PROCUREMENT DIAGRAM 1.3: -TRAFFIC DIAGRAM 2.0: ACCOUNTING INFORMATION SYSTEM DIAGRAM 2.1: ACCOUNTS RECEIVABLE DIAGRAM 2.2: PAYROLL AND LABOR DISTRIBUTION DIAGRAM 2.3: ACCOUNTS PAYABLE DIAGRAM 2.4: GENERAL LEDGER xvi 160 161 162 163 164 165 166 167 168 169 170 LIST OF SYMBOLS System Flow Chart Symbols Manual Operation Main Processing Function Punched Card Magnetic Tape Document or Report Direct Access File Storage 3) \_ D xvii ["flflafi”flfl]Terminal Batch Control Tape \N Communications Link ———) Informat ion Flow Offline Storage File ' Off Page Connector 0 On Page Connector LIST OF SYMBOLS cont'd. Data Flow Diagram Symbols Data Process Data Flow NV Data File Data Sink Entity-Relationship Diagram Symbols Entity Relationship . Key Attribute {D Non Key Attribute Generalization xviii CHAPTER I INTRODUCTION The integrated physical distribution management concept1 has received widespread approval within the business and academic communities. However, while the last decade has produced major breakthroughs in the understanding of the magnitude and importance of distribution costs, a growing amount of evidence indicates that the lack of adequate cost data. in the form. required for distribution management is preventing the integrated physical distribution management concept from reaching its full potential. Physical distribution is a multifunctional discipline which attempts to integrate a large number of different activities to meet a stated objective of cost minimization. To support these activities an accounting system which enables management to monitor routine operations and to integrate data for higher level decision purposes is required. A substantial amount of literature, however, 1"Physical distribution management is the term describing the integration of two or more activities for the purpose of planning, implementing and controlling the efficient flow of raw materials, in-process inventory and finished goods from point of origin to point of consumption. These activities may include, but are not limited to, customer service, demand forecasting, distribution. communications, inventory' control, material handling, order processing, parts and service support, plant and warehouse site selection, procurement, packaging, return goods handling, salvage and scrap disposal, traffic and transportation, and warehousing or storage." [NCPDM 1976] is critical of the ability of the existing accounting system to provide the necessary information for distribution management. For example, recent studies by ‘Lambert [1978], and Kearney, [1978] indicate that few firms presently know or have the type Of distribution cost information which enables confident strategic and operating decisions to be made. Strategic issues are often decided with faulty and misleading accounting data while Operational management has relied heavily on the use of averages and static control measures. Furthermore, while many impressive advances in the development of logistical models have occurred, the capability of the accounting system to provide the necessary data for these models appears to be lagging. 2 and the cost The importance of physical distribution associated with its activities suggests that increased attention be paid to the development of an accounting system which satisfies these input requirements. Consequently, the purpose of this dissertation is to extend and test a methodology for accounting for distribution activities. The methodology is based on the capture and retrieval of a wider variety of Operational and strategic data than is possible under traditional accounting systems. In addition the methodology permits the integration of the many diverse sets of information which characterize the physical distribution process but which previously have not been adequately integrated. 2 Physical distribution represents a staggering expenditure in this country» It has been estimated [Stewart and Mbrehouse, Background Until the last two decades, physical distribution had been managed as a set of fragmented and Often uncoordinated activities. Such fragmentation Often meant, for example, that transportation policies were carried out vdthout regard to subsequent inventory carrying charges. This lack of coordination generally resulted in higher total corporate costs. However, "the notion that a firm's total distribution costs could be reduced, customer service improved, and interdepartmental conflicts substantially reduced by the coordination Of distribution activities has emerged as an important concept. This concept has become known as the 'integrated physical distribution management'. [Lambert, Robeson and Stock, 1978, p. 74]. Recognition of the integrated nature of distribution activities has led to an expanded view of the physical distribution mission. This mission is to develop and maintain a distribution system that meets a stated corporate service level at a minimum cost. Achievement of such a goal requires two levels of integration: (1) integration of the physical distribution system with other corporate systems and (2) integration_ Of individual distribution activities. Operationalizing the first level requires 1978, p. 5] that in 1978, industry spent more than $200 billion on the transportation of products, over $125 billion storing them, more than $75 billion carrying them in inventory and approximately $20 billion in order processing, communication and administration. 3 standard be specified in terms of that a customer service timeliness and consistency and that the change in total enterprise cost with respect to a change in customer service be known. Operationalizing the second level requires the ability to minimize total distribution costs given a specified customer service level. For example, at a predetermined customer service level, total costs can be decreased if the increase in freight costs due to a policy of warehousing consolidation is less than the savings achieved by closing field warehouses. In short, the discipline of physical distribution has evolved over the years from an uncoordinated set of activities to the point where emphasis is now focused on a systems approach to managing a set of complex relationships. Sophisticated logistical models capable of a comprehensive analysis Of the total physical 4 distribution systems are availabLe. Customer service level and the total cost conceptS provide the conceptual vehicles for 3 Examples of customer service standards include: time from order placement to delivery; order cycle consistency; in stock availability; response time to variations in order size and assortment; delivery of order in satisfactory condition and a host of performance ratios relating to inventory availability. For' a more detailed investigation. see LaLonde and Zinszer [1976]. 4 See for example, the LREPS model deveIOped by Bowersox in Donald J. Bowersox, Qngistical. Management, [McMillan Publishing Co. Inc., New York, 1974], p. 395-409. 5 The concept of total cost was developed as a umasure of all expenditures required to complete the distribution mission. Flaks [1963] and LeKashman and Stolle [1965] provide a comprehensive treatment of the cost behavior patterns of specific distribution .activities and their impact (n1 total cost. measuring system performance. The theory behind the integrated distribution concept and the tools to make it operational appear to be in place. The Business Problem Effectively implementing the integrated distribution management concept requires a highly supportive accounting system which can capture, integrate and retrieve data for a wide variety of distribution users. Accounting systems, however, have not kept pace with the developments in physical distribution. Indeed, many writers contend that data required for decision making, control and trade-off analysis is either unavailable, unusable or distorted by financial accounting conventions. These problems appear to stem from two main sources. First, the development of distribution cost accounting has generally received low priority from corporate management resulting in classification schemes which do not fully isolate individual distribution activities. Without a more complete knowledge of the nature of the costs of these activities, it becomes difficult to make sound strategic and control decisions. For example, while a firm may be able to Obtain its total freight or warehousing cost, its accounting system may not have the information to determine its cost by delivery or its costs associated with receiving, storing and packing a given product or product line. Such deficiencies affect the firm's ability to capitalize on its most profitable routes, to choose its optimal facility configuration, and to determine how changes in one sector may affect the costs in another. Information contained in the chart of accounts is often classified in a manner which is not useful to the distribution decision maker, and the accounting system, with its emphasis on double entry, is seldom extended to incorporate the type of multidimensional data critical to the management of distribution activities. Second, the lack of adequate cost data and accounting involvement in distribution activities has led to the development of database systems which emphasize such items as shipment open order' files, inventory’ status files, forecasted. sales files .and product description files, Inn: which. contain ‘virtually' no information on distribution costs [House and Jackson, 1978]. Decision making and control then must rely on information which is predominantly non-cost oriented, which is based on financial accounting aggregations or which is gathered outside the system on an ad hoc and informal basis. These developments are regrettable for the following reasons: 1. The lack. of cost data severely limits the Iability' of management to carry out its planning and control duties within the distribution function. Similarly, since distribution cost data is a required input into higher level management models (e.g. adding or drOpping product lines), inability to accurately specify these costs increases the risk of inappropriate decisions at higher levels. 7 2. A variety of independent data gathering systems is unlikely to be effective in dealing with the numerous interactions which take place both between distribution and such systems as production and marketing and between the distribution components themselves. 3. Unnecessary and expensive duplications of effort exist when management cannot obtain the relevant information it requires from the database systems. In short, designing, implementing and Operating an effective distribution system requires a large number of information inputs for a wide variety of users. To a large extent, the cost information is derived either from accounting records utilizing traditional classifications and reports or from independent data gathering systems. Both approaches are unsatisfactory. In the first instance, data derived from accounting aggregations is seldom sufficiently detailed to satisfy the multiple demands of strategic and operating management. In the second, reliance on ad hoc data gathering systems leads to inefficiencies and overlapping of effort. [AAA, 1969, p. 52]. The Research Purpose Such problems represent a formidable research challenge. In particular they suggest a need for an approach to the accounting for distribution activities which provides a more complete capture Of the data underlying the relevant events affecting distribution Operations and which integrates this information with other corporate information systems. Consequently, the purpose Of this dissertation is to develop and test a generalized distribution information gathering methodology containing these features: 1. The methodology is based on an established accounting theory. 2. The methodology integrates the developments in accounting data modeling with those of database design approaches. 3. The methodology incorporates both non-cost and cost data in such a manner as to permit the generation of a variety of reports for a large group of users whose information needs may differ. The research relies on the events theory of accounting as first proposed by Sorter [Sorter, 1969]; a data model, the Entity- Relationship (E-R) Model, as outlined by (RENT [Chen, 1976] and extended by McCarthy [1979, 1982]; and the database design methodologies of Lum 35.31} [1979] and DeMarco [1979]. In effect, the purpose of the dissertation is to combine advances in data modeling with an acceptable accounting theory in order to provide timely and relevant information for distribution users. The Research Objectives The objectives of the research are three fold: Research Objective 1 Existing accounting systems have been subject to enough criticism to warrant an investigation into an alternative method for producing desired distribution. accounting information. Therefore, the first research objective is: 9 To extend the contributions of Sorter [1969], Chen [1976] and McCarthy [1982] into a generalized methodological framework for analyzing, capturing and retrieving the data needs of distribution management. Research Objective 2 Several areas of distribution management for which data has traditionally not been provided by the accounting system have been identified in the literature. In these cases, relevant data has either been missing from decision and control models or has been provided by an ad hoc data gathering system outside the existing accounting structure. Therefore the second research Objective is: To test the methodology using three examples of data insufficiency to determine if the methodology provides the desired information. Research Objective 3 The final research Objective is: To explore and identify avenues for future research. Scope and Limitations The research proposes a methodology for improving distribution cost information based on the events theory of accounting and on the design features of entity' and relationship sets. A. brief discussion Of some issues related to the methodology but not addressed by the dissertation follows. First, emphasis has been. placed oum umoo wcfiusuumwscms mo Noe mmma poacuo>mue .ooow mm zoom muusooua msam> 30H pom nozmwz .meuo>w co NONIQH qnma unmamum mucucm>cw cu umoo uwoumucfi am mcfiusacfi muomon .mumoo aofiuoseoaa co NmmIoa NHfiIm «mama muacum qu mnmfi xoo umoo mammaou HauOu mo qulwa ohma .Hm um mocoqma umoo mo meucmouwm m m< mmfimm mo Ommucoouom m m< mama nocua< mumoo :oHuanHuuwmm HOOszcm mumoo aofiusnwuumfin moonwanm mo moumawumm H mqm<fi 21 distribution management concept remains to be achieved. Accounting systems have not kept pace with the develOpments in physical distribution. Consequently, much of the data essential to distribution management is either collected in ad hoc fashion at different departmental levels, or captured by the financial accounting system resulting in levels of aggregation which are generally not useful for distribution uses. A review of the distribution literature reveals a large amount of criticism relating to existing accounting systems but little work addressing the problem of improving distribution systems accounting. Cost and management accountants have refined the process of factory cost accounting but have made little effort to extend this expertise to the distribution function.1 Wendell Stewart [1969, p. 18] recognized these problems several years ago. Perhaps the first and umst serious of these problems is the lack of adequate data. By this we mean that distribution costs in most companies are usually too gross and not available in the fine detail needed to conduct an accurate evaluation. of alternative distribution. methods and systems. Also, many of the costs that should be in view are hidden in vendor invoices and buried in other 1 The heavy emphasis on manufacturing accounting principles is evident in leading cost accounting books. For example, one of the most popular cost and managerial texts, Horngren [1982] does not mention distribution costs once in 934 pages. Altogether non-manufacturing concerns account for six pages. 22 cost centers such as manufacturing and marketing. Another aspect of this problem is the fact that information on sales volumes is frequently accumulated by production and warehousing locations, the source from which shipments are made rather than by market, that is, destination for which shipments are intended. Sawdy [1972, p. 1] indicated the nature of the problem in the opening sentence of his book. Many discussions about reducing distribution costs come to an abrupt halt because participants find they do not know what their distribution costs are at present. There is no point considering change if there is no knowledge of what it is that is to be changed. In the same year, Wilbur Wayman [Wayman, 1972, p. 33] focused on the central issue of accounting and the distribution function. If cost tradeoffs are at the heart of the logistics concept, then adequate cost information is at the heart of cost tradeoffs. Wayman criticized the inablility of the accounting system to provide the type of data useful to distribution management. Two of the earlier empirical studies support these contentions. Schiff's 1972 study pointed out the limitations of existing systems in producing meaningful accounting data. In the 14 companies surveyed, Schiff found that companies failed to capture physical distribution costs at the point of incurrence. He summarized this deficiency by noting that "in a significant number of companies, top management could not estimate total cost of distribution as a percentage of sales" [Schiff, 1972b, p. 3-1]. 23 A similar study was conducted in the United Kingdom in 1974 with similar results. The ifidtehead National Survey of Physical Distribution. Management noted that while the integrated distri- bution management concept had been accepted in principle, few firms had reached the implementation stage. Lack of adequate cost data was cited as the major problem [Bream and Galer, 1974]. Throughout the middle seventies, similar criticisms appeared. POpe [1976, [h 125] for example, noted the fOllowing deficiency. Figures from the financial accounts can generally provide little more than a beginning: there are often no weights or volumes and certainly no distances, while the breakdowns provided are usually insufficiently detailed for the purpose of distribution cost analysis. This can also apply to figures traditionally provided by the management accountant and factory accountant. Constantin, Anderson and Jerman [1977] found strong support for the hypothesis that current accounting practices were not adequate for decision making and control models in distribution. Shirley [1977] analyzed the current state of distribution accounting and concluded that greater research efforts were needed in solving the problems faced in distribution accounting. Lambert's [1978] study of the distribution channels decision of eighteen companies led him to conclude that decisions regarding the choice of channel and the evaluation of channel members would be substantially' improved if better cost and revenue data ‘were 24 available to the managers. Typical of the responses he received is the quote [1978, p. 88] which appears below. We are churning out information that isn't useful operationally. Sales by customer, sales by region would be useful. Profit by customer would be good to have. Transportation cost by customer would be interesting. It leaves me cold sometimes to say I wonder if we make money on sales to a location ...I don't know. Additional evidence of accounting data limitations was provided by the 1978 National Council of Physical Distribution Management Study entitled "Measuring Productivity in Physical Distribution." In a major survey of distribution accounting practices of North American firms, the study concluded that distribution accounting is still in an early stage of development. Only 5% of the firms surveyed had reached that point of evolution in their costing systems which were conducive to the implementation of the integrated distribution management concept. The report further noted the paucity of productivity data. [Stewart and Morehouse, 1978, p. 24] ...only a very small percentage of survey respondents have developed truly' meaningful productivity’ measurement systems for physical distribution. Thus there is a vast Opportunity to improve the measurement techniques being used by companies responding to our survey. Since these companies tend to already be among the most sophisticated, the Opportunity for improvement in U.S. industry is even more substantial. The need to collect and analyze large amounts of data in order to implement the integrated distribution concept necessarily means that cost should be an important ingredient in a company's database 25 systeuu. However, as illustrated by Christopher' and Ray [1976, preface] cost data has not been an important ingredient in database design. It has long been acknowledged that one of the main forces holding back the implementation of total distribution policies within companies has been the inadequacy of their cost database. If knowledge of costs is lacking then clearly it becomes impossible to identify potential trade- offs within the distribution system. The problem of missing costs in the database is also noted in the study by House and Jackson [1978, p. 63]. In general the survey indicated that computerized data bases for distribution maintained quantity data but not cost data. Warehousing costs, inventory carrying costs, stockout costs and packaging costs are not part of the typical data base. Finally, House and Karrenbauer [1978, p. 197] make the following observation with respect to data limitation in distribution modeling. ...the first overall limitation to large scale modeling is the fact that detailed data is mandatory. Unfortunately in many companies the information system is simply not set up to conveniently break out the kinds of relevant logistics data which are essential for a model. If the information is not available then a system has to be set up to gather it or special studies must be undertaken to obtain the necessary data. The Generalized Accounting Problem While the above comments are directed toward the data limitations faced by the distribution function, it should be noted that the phenomena just described are not unique to distribution. Similar problems can be found throughout most business organizations and to a large degree can be attributed to an 26 accounting system which, despite the significant impact of computer technology, is still rooted in double entry tradition. McCarthy [1980a, p. 628], for example, identifies four defects in the conventional accounting framework which impede the development of integrated accounting information systems. 1. Its dimensions are limited. Most accounting measurements are expressed in monetary terms: a practice that pre- cludes maintenance and use of puoductivity, performance, reliability, and other multidimensional data. 2. Its classification schemes are not always apprOpriate. The chart of accounts for a particular enterprise repre- sents all of the categories into which information concerning economic affairs may' be placed. This will often lead to data being left out or classified in a manner that hides its nature from non-accountants. 3. Its aggregation level for stored information is too high. Accounting data is used by a wide variety of decision makers, each needing differing amounts of quantity, aggregation and focus depending upon individual personalities, decision styles and conceptual structures. Therefore, information concerning economic events and objects should be kept in as elementary a form as possible to be aggregated by the eventual user. 4. Its degree of integration with the other functional areas of an enterprise is too restricted. Information concern- ing the same set of phenomena will often be maintained separately by accountants and non-accountants, thus leading to inconsistency and information gaps and overlaps. It is perhaps important not to leave the impression that the traditional accounting system has remained static over the past few decades or that it has been unaffected by technological develop- ments. Clearly this is not the case. The computer revolution has radically transformed the manner and speed with which accounting 27 information is processed, stored, retrieved and communicated. Authors such as Eaves [1966], Matthews [1967], Colantoni, Manes and Whinston [1971], Lieberman and Whinston [1975], and Haseman and Whinston [1976] are just some of those who have attempted to integrate theories of accounting with theories of information systems. Portions of this work are reviewed in the next section. Yet despite these advances, the basic tenets (n? the underlying accounting model have not changed in any significant way. The historical accounting emphasis on duality of double entry, on chart of account classification, on transactions which are primarily monetary in nature and on traditional aggregation rules has remained essentially unaffected by data model developments. It is these conventions whiCh are at time root of many of the problems indicated in the earlier quotes. McCrae recently cautioned that the distinction between changes in technology and changes in accounting systems must be recognized. The clear distinction between system. and technology' is important since many accountants suffer from the delusion that because they have changed the accounting technology they must automatically have affected dramatic changes in the accounting system. This is not so. The new computer technology provides a dramatic improvement in the speed of data processing and automatic control but this potential cannot be realized without affecting major alterations in the accounting system. In other words the accountant must develop more sophisticated accounting models to benefit from computer technology. [McCrae, 1976, p. 39] An example from distribution will reinforce this point. Implementing the integrated distribution management concept 28 requires a knowledge of how one functional area interacts with another. Specific examples of this include questions of trade-offs between transportation costs and warehousing and inventory carrying costs or between production costs and inventory carrying costs. However, the principles underlying current accounting procedures, whether manual or computerized, restrict this data integration effort. As McCarthy [1980a, p. 634] points out: This situation occurs because the double-entry’ form «if recording captures only selected characteristics of economic phenomena and then confounds the situation. by aggregating those characteristics over time and sections. Non-accountants interested in the same phenomena are faced with the dilemma of taking the accounting data as given or developing separate information systems. The next section reviews the development and application of an alternate accounting information model. This model is based on a disaggregate and multidimensional approach to the capture and maintenance of accounting data and has become known as the "events" accounting theory. Review of Events Accounting Literature and Its Early Applications In recent years a number of authors have supported the development of disaggregate and multidimensional approaches to the capture and maintenance of accounting data. Sorter [1969] and Johnson [1970] were the first to deveIOp the theory. Later authors attempted to Operationalize it by integrating the concept with different types of database systems. This section. reviews the 29 .a .Q .AmeH .zanwv .mcofiumuficssaou .mm. ecu mo 3OH>mm m unsoumzm mcfiucsooo< Oumwwuwmmmwa ocm HmconcoaHvfiufiaz: .%Luumuuz .m EmfiHHHB :oHumNHHmauoz mocoocoamoCH mama mCONuocsm wcwusuosuumom we cofiufichmQ wmmnmumo mucm>m mo cofiumuwcmwuo HmOwcoumuwwm wofiufifiwnmamo ommnmumm wsfinficmquIMHom wofiuwfiuoDomumzo ommnmumn oocfiwoaluomb ousuosuum uumm mouse muawwa< >oM wcwooo uco>m mumoocoo ommnmumn mo :oHuosoouucH Home: HOOfiumamcumz cofiumwwumw< manmmmwauom mo cofiuficfimmn mfiumufiuo :oHumoHMHuo> HmcoHum>uomno can ummoouom mmfism Hoseaumuoao ADOOLH osfim> mo mowmucm>omwfia mcfiucsooo< :muco>m= mm :Oumcwzz ocm :mEmmmm scuwcfisz can cmeuonoHA COOOCch new moon: .HCOucmHOU pounced powwow MOEHD< mmHofiuu< w:wuc:ooo< muco>m N mqmm no mo wcfiusuosuum < nebumhm cofiumauomcH one wcfiucaooo< mo zuom£fi may Cu nomouam< omwmfics < wcfiucsoou< mo snooze :mu:o>m: cm nausea zuowcH wcfiuasooo< memm on somouaq< :muco>m: q< MAHHH ”wousom mmmfi whom mmafi Huma Onma moaa m. m masoHe Hz.oaopumo Hz.mcamama owu ooo.s am ooH Rom .Om cam: mNH mamas: waa msHm> umHHon muovuo aofiumooq mamz Hooco> .oz Hovcm> mo .02 mucouo mucouo wwopoo< mmmuov< mausuom Ommnouzm _Ommcouam .uuaaom noono> mamz nooco> .oz pooco> hex >umafium mmaoauam uom ODHm> auspaauu< 47 The Chen data modeling methodology can be summarized in the following steps. 1) Identification of the entity and relationship sets in the system being modeled. 2) Identification of the semantic correspondence between the sets through an entity-relationship diagram. 3) Identification of the attributes and values associated with each set; and 4) Organization of data into relational tables. Using the Chen methodology as a base, McCarthy [1979] developed an accounting data model. His E-R model differs from the conventional accounting model in two respects. First, it allows the admission of 21 wider variety of elements into the system. Second, the traditional chart of account classifications and ledger summaries are replaced by a number of entity-relationship sets. Such a scheme permits the development of a database system in a manner superior to the conventional accounting approach whose "artifacts" were the subject of earlier criticism by Everest and Weber. Thus McCarthy's E-R accounting system was the first events system that was not predicated on prevailing accounting practice. McCarthy [1982] has since extended his model to an REA model (resources, events, agents) which serves as a generalized framework for accounting systems in a shared data environment. The review of Sorter's events accounting theory and the data model developments undertaken to Operationalize that theory is now 48 complete. The next section briefly reviews the theoretical aspects of database design as certain of these features will be incorporated in the next chapter. Review of Database Design Literature Relevant to the Research Recently the process of designing a database to support numerous and complex applications has received increasing attention from theorists. While this research is still somewhat embryonic, four separate but dependent areas of the design process appear to have emerged [Lum E£.§lf’ 1979]. 1. Requirements Analysis 2. Information Analysis and Definition 3. Implementation Design 4. Physical Design All four represent critical steps in achieving the attributes of a sound information system as outlined earlier by Everest. However, the last two refer more closely to the actual management of the database and to particular software packages. Since these procedures primarily involve machine efficiency criteria and because the model to be employed during the research can be used with most of the existing physical systems, steps 3 and 4 are beyond the scope of the dissertation. Steps 1 and 2 have a direct bearing on the dissertation. Their role is now highlighted. 49 Requirements Analysis Identification of corporate information needs has always been of critical importance to the successful implementation of an information system. As the trend toward integrated database systems becomes more pronounced, the importance of this step of the design process will continue to increase. In an environment where data is viewed as a corporate resource and where data has a multi- user orientation, no design is valid unless the information needs associated with the end users are well defined. IAchieving an acceptable level. of information system: performance is :1 complex task and depends heavily on the success with which the requirement analysis phase is carried out. This section briefly outlines the role of requirements analysis in the overall design of the database system and describes how these requirements would normally be Obtained in the field. Examined in this section are the contributions of Lum 35.31} [1979] and DeMarco [1979]. Requirements analysis can be simply defined as a procedure which attempts to determine the present and future information needs of the user community' where the user can range from an invoice processing clerk to a tOp level manager responsible for the planning and execution of the strategic plan. While this aspect of 3 the design phase has been described by many different writers, an excellent description of the procedural aspects associated with 3 See for example, Taggart and Tharp [1977], Carter 53; El) [1975] and Yao, Navathe and Weldon [1978]. 50 requirements analysis is provided by Lum §£_§13 [1979]. Essential- ly the process involves all levels of management and has the intention of pmoviding information about data and processes which can in: used in. subsequent stages of database design. Diagramatically, this activity and its interaction with later phases of database representation is illustrated in Figure 6. First, top management provides the overall scope of the information system. Included at this level are such corporate guidelines as (l) objectives and goals, (2) resources available to achieve the desired outcome, (3) strategic alternatives, (4) effects of changes on current business operations, and (5) operating procedures and constraints. Sollenberger [1971] outlined the importance of accurately specifying these guidelines in the successful develOpment of an) information system. Once obtained, they provide direction as to the relative importance Of individual business functions and act as a consistency check on information needs obtained at lower levels of management. Second, middle management provides information about data required for yflanning and control functions as well as information regarding aspects of data security, reliability and response times. Third, Operations management provide descriptions of the data and processes pertinent to their own applications. For example, an accountant might define his view of the data needed to complete a warehouse payroll while the warehouse manager would provide a different view of the data required to measure labor efficiency. A complete definition of a 51 TOP MIDDLE OPERATIONS DATABASE - MANAGEMENT MANAGEMENT MANAGEMENT ADMINISTRATOR INPUT INPUT INPUT INPUT R E Q U CORPORATE REQUIREMENTS I R A ANALYSIS E N ' M A I I .. CORP RATE INFORMATION PROC SSING N Y CONST INTS REQUI EMENTS REQU REMENTS T S S I S VIEW MODELING I N AND MODIFICATION F ‘ O I I I R MANA ERIAL APPL CATION ACCE S REQUIREMENTS M VIEW VIEWS OF APPLICATIONS A A T N I D ._ 0 VIEW MODELING N D E AND INTEGRATION A F r N I 1 AN ENTE PRISE L I VI W Y T j s. -— I O S N IMPLEMENTATION AND PHYSICAL DATABASE DESIGN STAGES Source: Adapted from Vincent Lum 35.313, "1978 New Orleans Database WorkshOp Report," Research Report RJ2554 (IBM Research Laboratories, San Jose, CA., July, 1979, p. 108). FIGURE 6 Diagram Illustrating Sequence Of Database Design Procedures 52 data object includes a description of its name, size, number of occurrences, reliability, security and relationships with other data. Similarly, a complete description of a pmocess requires a knowledge of how data is retrieved, updated, deleted and used in specific applications. 11m: gt Elf [1977] provide further information regarding the procedures by which information requirements are obtained at different levels of management. Whereas interviews, questionnaires and narrative documentation are traditional vehicles employed for obtaining knowledge an: the top and middle management levels, the procedure for obtaining information at the operations level is more highly structured. The authors however, provide only minimal guidance as to the form such structuring should take. The structured analysis procedures advocated by DeMarco [1979] provide a more informative way of collecting these requirements. DeMarco begins with the premise that classical analysis has concentrated (n1 detailing the wmrkings of the Operation from the point of view of the user rather than concentrating on detailing the workings of the operation from the point of view of the data: He defines this latter orientation of following the data through the Operation as "interviewing the data." [1979, p. 49] Since this view has the advantage of allowing a complete picture of the current operating procedure to be developed, he describes the technique as "more productive than any other single interview." [1979, p. 49] Figure 7 illustrates those procedures which are | W I UKEH' I 'nuwurds' l w u . I may ”I" ’ weesm' | - IRONMENT I "‘4“ ‘ I l DEENE I WON. (Lavenw EamukuiadT I “we-“4 9:2“ I “343% I I I " Ciqgnnnns I 0;;qu new Data we. CAL. I %:3:7: $Y;TEM 0mm I I I q,” a“ tum/e1 I Mafiahbn Cl: fihkfibflfly I £45525 " I I I I I l l. I . L REQUIREMENTS ANALYSIS INFORMATION ANALYSIS ] STAGE AND DEFINITION STAGE Source: Adapted from, Tom DeMarco, Structured Analysis and System Specification, Prentice-Hall, Englewood Cliffs, New Jersey, 1979, p. 26. FIGURE 7 Components of Structured Analysis 54 concerned with the requirements analysis and information analysis and definition stages. Circles or 'bubbles' are used in the figure to represent processes while the arrow or 'arcs' represent data flows. The rectangular box or 'sink' denotes a source of informatdxnl. DeMarco's approach to obtaining information..about user requirements is now summarized. 1. In conjunction with users, a verifiable model of the current business environment is constructed. The ‘user supplies sufficient information about data and processes to allow a current physical data flow diagram to be constructed. The diagram is intended to be a portrayal of the existing information system and contains the kind of detail about such familiar objects as departments, people's names, document names and form numbers so as to be understood by the user. 2. The physical checkpoints in (l) are eliminated and a current logical equivalent (M? the information system is derived. This model provides a description of the data and processes currently utilized without including the departments or people who use them. 3. Changes that are to be made in the system are identified. These changes may result from surveying the users to determine which needs are not currently being satisfied. These data needs are described in a report such as a feasibility document and converted to a logical form. The final product of requirements analysis is a provision of (1) information concerning corporate constraints imposed on data usage and (2) information concerning the data and processes required by managers to support their decision making activities. The above constitutes a brief description of the purpose, importance and nature of requirements analysis. Procedurally, the entire process is a highly labor intensive and time consuming operation. In addition, substantial trial, error and revision is 55 normally required before reliable results can be achieved. Although a few data models exist which are designed to document individual requirements (such as the PSL/PSA model developed by Teichroew and Hershey, 1977), none are adequate for communicating directly with the user because of their high degree of abstraction. Thus, while this phase of database design has achieved recognition as an integral part of the information system development process, it is still fraught with communication problems and imprecision. In 'view (n? these field problems, researchers interested in data models have employed alternative methodologies for accomplishing the requirements analysis stage. One such methodology is to develop a case study and specify the constraints and information needs for a hypothetical firm. Such an approach has found widespread acceptance in the academic literature and is exemplified by the Bubenko [1977] and Teorey and Fry [1980] approaches to requirement analysis formulation in the design of database systems. Once the information from the requirements analysis phase has been obtained database design proceeds to the information analysis and definition phase. This is briefly discussed in the next section. Information Analysis and Definition Information analysis and definition is concerned with the conceptual modeling of the database as a formal representation of 56 the information obtained from requirements analysis. Both Figure 6 and 7 portray this stage. However, the following brief description is oriented to the Lum e_t_ El. model (Figure 6). As indicated in this diagram, information analysis and definition consists of two phases. View Modeling and Modification involves transforming the information anui processing requirements obtained earlier into an underlying conceptual model. The output from this phase of database design includes the following views. 1. Managerial views representing models of the corporate information requirements. 2. Application views representing the information required to support individual applications. 3. Access requirement views representing both corporate and application processing requirements. View Analysis and Integration analyzes the individual views and integrates them into an overall enterprise view. Inconsist- encies and differences in perception which may exist among users are eliminated at this point. The outcome of view integration is a consistent enterprise view of data which is independent of the specific database management system used to process the data. Information analysis and definition is based on an underlying high level conceptual model. These models, such as the E-R model, serve two purposes. First, they provide the framework for understanding the information system requirements by describing its entities, relationships and attributes. Second, they provide the 57 means by which such information is transferred to the implementation stage of design. Further descriptions of the procedures and models used during information analysis can be found in Lum 33 El' [1979], Yao, Navathe and Weldon [1978] and Fry 33.31} [1978]. Summary A review of the literature underlying this dissertation is now complete. Consideration has been given to these areas. 1. The nature of the distribution function and the data inadequacy problems it faces under conventional accounting doctrine. 2. An alternative accounting theory, based on an events approach, which is more consistent with contemporary data environments. 3. The first attempts to Operationalize the events approach and the accounting artifact problems which arose with the use of first generation data models. 4. A second generation data model approach which overcomes the accounting artifact problem. This approach allows the accountant and non-accountant to work together in modeling their particular financial and managerial accounting data needs in a manner consistent with the theory of database design. 5. The aspects of database design that are relevant to this research. It has been argued that the entity-relationship data modeling approach to accounting system develOpment overcomes a number of limitations inherent in models which rely on traditional accounting constructs. To date, however, applications of the approach to accounting have been limited to McCarthy's E-R construction of a 58 small retail enterprise [McCarthy, 1979]. Consequently, this dissertation represents an extension to such work by exploring and modeling the data needs of the distribution function. The research should prove fruitful since it covers many more managerially related issues than. could ‘McCarthy's financially-oriented retail representation. The research methodology is discussed in the next section. 59 References American Accounting Association (1969), "Report of Committee on Managerial Decision Models, "The Accounting Review, (Supplement 1969), pp. 43-76. Bream, R. E. and R. Galer (1974), A National Survey of Physical Distribution Management, (Whitehead and Partners, 1974). Bubenko, J. A. (1977), "IAM: An Inferential Abstract Modeling Approach to design of Conceptual Schema, "ACM-SIGMOD International Conference on Management of Data, (August, 1977), pp. 62473} Carter, D. M., H. L. Gibson and R. A. Rademacher (1975), A Study of Critical Factors in Management Information Systems for the U.S. Air Force, (Colorado State University, 1975). Chen, P. (1976), "The Entity Relationship Mbdel--Toward a Unified View of Data, "ACM Transactions on Data Base Systems, (March 1976), pp. 9-36. Christopher, M. and D. Ray (1976), Costing in Distribution, (MCB Books, West Yorkshire, England, 1976). CODASYL Programming Language Committee (1971), Data Base Task Group Report (Association for Computing Machinery, 1971). Codd, E. F. (1970), "A Relational Mbdel of'Data for Large Shared Data Banks," Communications of the ACM (June 1970), pp. 377- 387. (l972a), "Further Normalization of the Data Base Relational Model," in R. Rustin, ed., Data Base Systems (Prentice-Hall, (1972b), ”Relational Completeness of Data Base Sublanguages,” 3H1 R° Rustin, ed., Data Base Systems (Prentice-Hall, 1972), pp. 65-98. Colantoni, C. S., R. P. Manes and A. B. Whinston (1971), "A Unified Approach to the Theory of Accounting and Information Systems,” The Accountinijeview, (January 1971), pp. 90-102. Constantin, J. A., R. D. Anderson and R. E. Jerman (1977), "View of Physical Distribution Managers," Business Horizons, (April, 1977), pp. 82-86. Cox, R., quoted in R. Moyer (1972), A Social Pegspective, John Wiley and Sons, (1972), p. 52. DeMarco, T. (1979), Structured Analysis and System flecification (Prentice-Hall, Inc., 1979). 60 Drucker, P. (1962), "The Economy's Dark Continent," Fortune, (April 1962). Eaves, B. C. (1966), "Operational Axiomatic Accounting Mechanics," The Accounting Review, (July 1966), pp. 426-442. Everest, G. C. and R. Weber (1977), "A Relational Approach to Accounting Mbdels," The AccountingyReview, (April 1977), pp. 340-359. Everest, G. C. (1974), "The Objectives of Data Base Management" in Julius T. Tou, ed., Information Systems: COINS IV (Plenum, 1974), pp. 1-35. Fry, J. P., T. J. Teorey, D. A. DeSmith and L. B. Oberlander (1978), Survey of State of the Art Database Administration Tools: Survey Results and Evaluation, Technical Report, DSRG 78 DE 14.2 Division of Research, Graduate School of Business Administration, University of Michigan, Ann Arbor, MI., 1978. Haseman, W. D. and A, B. Whinston (1976), "Design. of [A Multidimensional Accounting System{" The Accounting Review, (January 1976), pp. 65-79. (1977), Introduction to Data Management (Richard D. Irwin, 1977). Horngren, C. T. (1982), Cost Accounting: A Managerial Emphasis, 5th ed., (Prentice-Hall, 1977). House, R. J. and B.R. Jackson (1978), "Trends in Computer Applications: A Survey,” in Contemporary; Physical Distribution by J. C. Johnson, (Petroleum Publishing Company, 1977678), pp. 62-66. House, R, J. and. J. J. Karranbauer (1978), ”Logistics Systems Modelling,” International Journal of Physical Distribution and Materials Management, Vol. 8, No. 4, (1978), pp. 189-199. Ijiri, Y. (1967), The Foundation of Accounting Measurement (Prentice-Hall, 1967). Johnson, 0. (1970), ”Toward an 'Events' Theory of Accounting," The AccountigggReview (October 1970), pp. 641-53. Kearney, A. T. (1978), Measuring Productivity in Physical Distribution, (National Council of Physical Distribution Management, Chicago, IL. 1978). LaLonde, B. J., J. R. Grabner and J. F. Robeson (1970), ”Integrated Distribution Systems: A» Management Perspective,” International Journal of Physical Distribution, (October, 1970), pp. 133-139. 61 Lambert, D. M. and J.R. Stock (1982), Strategic Phj_sica1 Distribution Manpgement, (Richard D. Irwin, Homewood, IL), 1982. Lambert, D. M. (1976), The DevelOpment of an Inventory Carrying Costing Methodologg A Study of the Costs Asociated with Holding Inventory, (Chicago, National Council of Physical Distribution Management, 1976). (1978), The Distribution Channels Decision, (New York: The National Association of Accountants and Hamilton, Ontario: The Society of Management Accountants of Canada, 1978). , J. F. Robeson and J. R. Stock (1978), "An Appraisal of the Integrated Physical Distribution Management Concept,” International Journal of Physical Distribution and Materials Management, Volume 9, Number 1, (1978), p. 74. Lieberman, A. Z. and A. B. Whinston (1975), ”A Structuring of An Events-Accounting Information System," The Accounting Review, (April 1975), pp. 246-258. Lum, V., S. Ghosh, M. Scholnick, D. Jefferson, 3. Su, J. Fry, T. Teorey and B. Yao (1979), "1978 New Orleans Data Base Design Workshop Report,” Research Report RJ2554 (IBM Research Laboratories, San Jose, CA, July, 1979). Matthews, R. L. (1967), "A Computer Programming Approach to the Design of Accounting Systems,” ABACUS (December 1967), pp. McCarthy, W. E. (1979), "An Entity-Relationship View of Accounting Models,” The AccountingReview, (October 1979), pp.667-686. (1980a), "Construction and Use of Integrated Accounting Systems with Entity-Relationship Modeling," in P. P. Chen, ed., Entity-Relationship Approach to Systems Analysis and Design, (North Holland Publishing Company, 1980), pp. 625- 637. (l980b), "Multidimensional and Disaggregate Accounting Systems: A Review of the 'Event' Accounting Literature," MAS Communication (July, 1981), pp. 7-13. (1982), ”The REA Accounting Model: A Generalized Framework for Accounting Systems in a Shared Data Environment,” The AccountingReview, (July, 1982), pp. 554-578. McCrae, T. W. (1976), Computers and Accounting, (London: John Wiley and Sons, Ltd., 1976). Pope, A. L. (1976) "The Concept and Cost Elements of Physical Distribution” in Costing in Distribution, edited by M. Christopher and D. Ray (MCB Books, 1976), pp. 109-136. 62 Ray, D. (1975), "Distribution Costing - The Current State of the Art," International Journal of Physical Distribution, Vol. 6, Sawdy, L. C. (1972), The Economics of Distribution (John Wiley and Sons, 1972). Schiff, M. (1972a), "Physical Distribution: a Cost Analysis," Management Accounting, (February 1972), pp. 48-50. (1972b), Accounting and Control in Physical Distribuion Management, (Chicago, National Council of Physical Distribution Management, 1972). Shirley, R. E. (1977), "Accounting Analysis of Distribution Activities--A Critique," International Journal of Physical Distribution, Vol. 7, No. 5, (1977), pp. 275-282. Sollenberger, H. M. (1971), Management Control of Information Systems Development, New York: (National Association of Accountants, 1971). Sorter, G. (1969), "An 'Events' Approach to Basic Accounting Theory, The Accounting Review, (January 1969), pp.12-l9. Stevenson, R. (1977), "Managing Physical Distribution," The CPA Journal, (May, 1977) p. 74-79. Stewart, W. M5 (1969), "P.D. Revisited" Proceedings of the NCPDM Fall Meeting, (Chicago, National Council of Physical Distribution Management, 1969). (1974), "Don't be Among the 99 44/100% of Companies Who Don't Know Their Distribution Costs" Handling and Shipping, (Presidential Issue, 1974), pp. 31-97. and J.E. Morehouse (1978), "Improving Productivity in Physical Distribution: IA $40 Billion goldmine," in Proceedings NCPDM Annual Meeting, (Chicago, National Council of Physical Distribution Management, 1978), pp. 1-33. Taggart, Jr., W.M., and M.0. Tharp (1977), "A Survey of Information Requirements Analysis Techniques," Computing Surveys (December 1977), pp. 273-90. Tavernier, G. (1975), "Controlling the Soaring Cost of Distribution," International Management, (August, 1975), pp. 11-16 a Teichroew, T.J., and E.A. Hershey (1977), "PSL/PSA: A Computer Aided Technique for Structured Documentation and Analysis of Information Processing Systems," IEEE Transactions Software Engineering, (SE-3, 1 1977), pp. 41-48. 63 Teorey, T.J., and J.F. Fry (1980), "The Logical Record Access Approach to Data Base Design," Computing Surveys, Vol. 12, No. 2, (June 1980), pp. 179-211. Wayman, W. (1972), "Harnessing the Corporate Accounting System for Physical Distribution Cost Information,” Distribution System Costing: Concepts and Procedures (James R. Riley, Symposium on Business Logistics), April 1972, pp. 31-46. Willis, R. (1977), Physical Distribution Management, (New Jersey, Noyes Data Corporation, 1977). Yao, S. B., D.L. Navathe and J.L. Weldon (1978), "An Integrated Approach to Logical Database Design," in New York Symposium on Database Design, (Graduate School of Business Administration, N.Y., N.Y., May 1978), pp. 1-14. CHAPTER III RESEARCH DESIGN The events theory of accounting and the entity-relationship data modeling approach represent a conceptually sound methodology for satisfying distribution, financial and managerial data requirements. Chapter III describes the research. design that allows such an accounting data model for distribution to be developed. Specifically the design accomplishes these steps. 1. Constructs a realistic example of a company's distribution operations. 2. Incorporates three examples from the literature where distribution data needs are not being satisfied by the traditional accounting information system. 3. Describes the existing flow of information through the distribution system. 4. Develops a logical view of both the existing procedures and new data requirements which together provide information about the data and processes to be incorporated in the entity-relationship model. In essence, the research design used in this study represents a particular form of the requirements analysis phase as outlined by Lum £5 31. [1979] and DeMarco [1979]. The general design and the adaptation on which the research is based is discussed next. 64 65 Following this, the research situation is described, and the new and existing user requirements are documented. Methodological Overview As indicated in the previous chapter, a critical activity in the design. of an information system is to determine the user requirements and to express these in such a way that they can be utilized by subsequent phases of the design process. The methodology for analyzing and refining the information requirements in this study represents a combination of the methodologies noted in Chapter II; namely the structured analysis methodology as outlined by DeMarco [1979] and the database design methodology as outlined by Lum e_p El' [1979]. Figures 8 and 9 portray these combined methodologies and are now described in more detail. Figure 8 is a data flow diagram (DFD) which illustrates a six step project life cycle design procedure. The diagram is a high level overview of a systematic approach which begins with the analysis of users needs and ends in the implementation of the new system. Each of the processes in Figure 8 can be further ”leveled" or decomposed to provide additional detail. Figure 9(a) and 9(b) are leveled DFD's which show in more detail how some of these transformations take place. For example, it can be seen that the process of structured analysis (which is labeled as bubble #2 in Figure 8) can be broken down into seven subprocesses (which are labeled as bubbles #2.1 through 2.7 in Figure 9b). It can also be FIGURE 8 Project Life Cycle Source: Adapted from Tom DeMarco, Structured Analysis and System Specification, Prentice-Hall, 1979. 67 FIGURE 9(a) Survey Data Flow Diagram 2:! awumry CLHZEIflwT user I ' IIRCAfllmd nurdrenlnhb :HVuohm eguu¢k;::L\ FIGURE 9(b) Structured Analysis Data Flow Diagram Source: Adapted From Tom DeMarco, Structured Analysis and Systems Specification, Prentice-Hall, 1979 68 seen that when a process is leveled, its corresponding diagrams "balance” in the sense that net inputs and outputs remain the same between levels. Use of this methodology leads to a number of desirable features. First, it overcomes the following defects that have been ascribed to more traditional analysis schemes (DeMarco [1979], Gene and Sarson [1979]. 1. Inadequate methods for describing procedures. 2. Non-maintainable narrative approaches in) system documentation. 3. Inadequate communication between users and analysts. 4. Failure to derive usable models at early stages of system development. 5. Inappropriate uses of flowcharts which commit users to specific physical implementations, and 6. Inadequate procedures for determining user preferences and tradeoffs for data. Second, the strengths of both the structured analysis and data base design approaches are incorporated into a single methodology resulting in a highly pragmatic set of procedures rooted in a solid theoretical foundation. Third, the methodology emphasizes the design and use of logical models thereby focusing on the question of what the information system should produce rather than on the question of how the system should be implemented. Such a separation allows the 69 accountant and designer to concentrate on analyzing user requirements and then model those requirements in a manner which can logically be passed on to later stages of design work. Methodology Used to Obtain and Analyze Information Requirements in This Study Figures 8 and 9 have been constructed with field applications in mind. The procedures outlined in these diagrams can be modified to permit the development of a specific research design for distribution which differs from a field study in the following way. Instead of interviewing users in an actual distribution setting, (as would take place in Figure 9(a)) the survey phase of Figures 8 and 9(a) can be replaced by a literature review to determine those areas for which needed data has traditionally not been supplied by the accounting system. Furthermore, the structured analysis phase (Bubble #2 in Figure 8) which begins with the study of the current environment (Bubble 2.1 in Figure 9(b)) can be modified to include a realistic case example of a distribution operation. Such an approach is similar to the hypothetical case study approaches used by Bowersox [1973], Bubenko [1977] and Teorey and Fry [1980]. These modifications are shown in Figure 10. Bubbles 1.1, 1.2, 2.1 and 2.2 are concerned with determining new and existing user needs and are collectively called the requirements analysis stage. This is the subject of the present chapter. Bubble 2.3 is concerned with deriving a fbrmal representation of the information elements 70 obtained from requirements analysis. This is accomplished with the use of Chen's E-R model and is the subject of Chapter 4 (Bubbles with dotted lines indicate those processes which correSpond to Figure 9(b) but which are only briefly touched on in this dissertation). Details of the requirements analysis phase are now described. New Requirements Analysis Requirements analysis is the determination of both future and present information needs of the user community. In this study, the new information requirements are derived in a research setting as opposed to a: field study of actual users. Prior to outlining the research procedures, it is useful to contrast the differences between the two (i.e. between Bubbles 1.1 and 1.2 of Figure 10 and Bubbles 1.1 and 1.2 of Figure 9(a)). These differences are not in the methodology per se but in the techniques used to obtain user requirements. New requirements analysis in the field involves selecting between requests which may be motivated by such factors as: l. A desire for improved response time such as providing up to the minute information on inventory levels. 2. The imposition of a new reporting requirement such as the recent FASB and SEC requirement for replacement cost data for inventory and fixed assets. 3. A desire for the development of a new system to provide additional decision making or problem solving capabilities 71 for such issues as the product continuance decision or delivery vehicle routing. In each case, the requests must be subjected to an initial justification. One such technique (IRACIS) involves an overview evaluation of a request's increased Eevenue, ayoided post or improved eervice potential. [Gane and Sarson, p. 156]. If these initial estimates are favorable, a detailed study (Bubble #1.2 of Figure 9(a) is carried out. If the detailed analysis also provides promising results, a feasibility document is prepared which includes the following: 1. A comprehensive description of the affected user community. 2. A statement of benefits associated with the change request; and 3. A precise statement of objectives that the new system is to satisfy including a specific enumeration of the new decision making or control processes to be supported and the data needs of each. The methodology used in the dissertation is similar to that described above except that the literature is used to identify new information requirements. Further, while a justification for each of the areas identified is provided, no cost benefit analysis of the system changes implied in the research can be meaningfully undertaken. 72 uhsuflthlcmy S a, : a 2 . 4.\ spao‘vficahbn ”man—Isa \ owns-finkuwnuul M‘LYQ‘ l W “m7!“ \ / mu» -— / 499+ me: In W new :éaJ 44;:q90u: cfiaanuN\ CHM ucaJ dqfin-Houa “'9 ' «late. eknnon+ l du'cd'tomy I WHOM "laufipfibn oftauwuw? ‘*”“‘Ph°“’ -—-r M 519% via mama... and 2 '7 67 g ( PM‘E- acumen-Ion, I \ / \ / W.M.... I Aknounkl Requirements Analysis- CH.III Information Analysis and Definition- CH.IV FIGURE 10 Diagram Illustration Components of Research Design 73 In this study, the new information requirements have been identified in the literature as examples where the unavailablility of accounting data has prevented certain key distribution control and decision needs from being effectively satisfied. Three problem areas have been selected which fall roughly into Anthony's [1970] decision framework of operational control, management control and strategic planning. The three areas are: 1. Accounting for Warehouse Labor Productivity 2. Choosing the Least Cost Distribution System 3. Warehouse Site Selection Decisions In addition to the above, the research intends to show how distribution financial reporting requirements can be achieved from the same data model that provides information for other applications. Since warehouse labor statistics are a necessary input for financial reporting purposes, the information requirements of the payroll function are also defined. The data requirements associated with each of these examples are obtained by producing a concise statement of management information. needs ‘which. are supported by' the literature. The result of this part of the design process is a number of independent views regarding data requirements. Once specified the new requirements are passed on to the information analysis and definition phase (Bubble #2.3 of Figure 10). 74 ‘Existing Reguirements.Analysis Figure 10 indicates how existing information needs are defined for purposes of this study. Once again, these differ from the methods used in Figure 8 land 9(a) not in substance but in procedure. Based on information in the distribution literature as well as on the author's own experience in the field, a description of a realistic set of distribution operations is first constructed. Contained. in. the description. are *various company' profiles (providing such information as size, product offerings and distribution linkages) and specific Operating procedures and constraints. Accompanying the description is a systems flowchart specifying the flow of data involved in each aspect of the distribution process and the linkage of this data to the accounting system. The symbolic definitions found in the systems flowchart are similar to those used by Cushing [1982] while the descriptions of data flows and file compositions are. constructed using the DeMarco [1979] procedures. The product of this part of the research design is a portrayal of how the firm currently carries out its distribution mission; a portrayal which includes: (1) the data it is currently collecting, (2) the processes it is currently using and (3) the reports it is currently generating. Included in the above descriptions are a number of identifiers which aid the distribution manager in understanding the information process. For example, department codes, individual's names, document numbers and processing symbols are incorporated to enhance 75 the interpretability of a narrative or systems flow diagram. Once this portrayal is complete however, the identifiers can be removed and a new data flow diagram constructed. This step results in a logical view! of the current information process which. is unencumbered by nonessential descriptors. Together with the statement of new requirements, the logical data flow diagram is then passed along to information analysis and definition (Bubble #2.3 of Figure 10). This sets the stage for the analysis in Chapter' 4 'which (1) models the individual managerial and application views and (2) integrates these views into a more global enterprise view' capable of providing the required distribution information. Advantages Associated with Research Desigp Several advantages accrue to this type of research design. First, the problems associated with studying the existing environment and collecting actual information requirements in a field study are alleviated. Second, the time required to complete the study becomes reasonable.1 Third, since all assumptions about the operating environment and information requirements are contained in the design, iterations not central to the methodological development are avoided. Finally, despite the 1 DeMarco [1979, p. 35] notes that the study of the current environment itself represents, "30 percent of the total 76 apparent abstraction from reality, the research design provides a situation which may be more generalizable than an actual field study. This is due to (1) a description of the distribution environment which is general enough to be familiar to most firms and (2) a set of problems which. have been identified in the literature as issues which affect a majority of firms who must distribute products. The next section describes the research situation. The Research Situation Bowersox ep 21, [1973, p. 25] note that the general nature of the physical distribution system model is that of a manufacturing firm which produces and distributes on a national basis through either company facilities or marketing intermediaries. This section provides an analysis of the research situation by describing a particular physical distribution network. which is consistent.with this general model. HP company is a medium size manufacturing firm headquartered in the midwest. Each of its production facilities (PF) produces a variety of products which are initially stored in warehouses (PFW) adjacent to the production facilities. Subsequently, the goods are transported to either regional distribution centers (RDC) or local analysis phase manpower” and Keen and Scott MOrton [1978, p. 185] note that obtaining information from users regarding key issues, overall objectives and available resources may take up to 70% of the elapsed time of a project. 77 distribution centers (LDC) situated throughout predefined areas in the United States. In addition, the company purchases a portion of its product offerings from outside vendors and stores this inventory at RDC and LDC locations. Each distribution center receives orders from and transports products to customers in its own geographical location. In cases of stock outs at the LDC level, the customer may be serviced by the RDC in that area. Merchandise may be transported between warehouse or distribution center locations and between such locations and customers in several ways. Vendor shipments to RDC's and LDC's are accomplished by common carrier; shipments to RDC's and LDC's from PFW's are delivered by common carrier or company truck; and RDC and LDC shipments to customers are made either by common carrier, company truck or United Postal Service (UPS). Figure 11 illustrates the distribution configuration. Further information concerning the nature by which the firm conducts its business is shown by the company profiles listed in Table III. Together, the illustrations portray a national company offering a wide range of products for a large group of customers through a distribution network linked by several transportation modes. The next sections are concerned with determining the needs of the HP user community with respect to both new and existing information. As illustrated in Figure 10, new requirements are 78 AREA 1 Key: 1 AREA 2 r AREA 3 —.‘ Flow of Product PF's and PEW's - Production facilities and production facility warehouses produce and store products for RDC's and LDC's located in one or more geographical areas. PDC's - Regional distribution centers service all customers assigned to the RDC region. May also service LDC customers in case of stock out. LDC's - Local distribution centers service only customers in LDC geographical locations. C - Customers V - Vendors AREA- Defined by management as a particular grouping of RDC's, LDC's and C's. FIGURE 11 Diagram Illustrating Flow of Product For a Physical Distribution Network 79 TABLE 3 Table Describing Various Profiles of the Company's Business Operations Headquarters Production Facilities Production Facility Warehouses Regional Distribution Centers Local Distribution Centers Transportation Mode Vendors Midwest United States Located throughout the U.S. Produce partial line of products. Output initially goes to warehouse adjacent to production facilities. Located adjacent to each production facility. Store partial line of products. Service RDC's and LDC's in one or more areas. Located in or close to major metropolitan centers throughout the U.S. Company owned or leased. Predominantly manual receiving and shipping operations. Service customers in RDC regions and may also service LDC customers in case of stock outs at LDC's. Store full line of company products. Located in or close to smaller communities throughout the U.S. Company owned or leased. Predominantly manual receiving and shipping Operations. Service customers in an LDC geographical location. Store full line of company products. Common carrier. Company owned vehicle. UPS Located throughout United States. Products are shipped directly to RDC's and LDC's. TABLE 3 cont'd. Products and Customers Communications Mode and Information System 80 Multiproduct lines each with one or more models offered to a varied customer clientele. Described in section entitled "Existing Requirements Analysis" 81 documented from the literature and are then passed on to subsequent stages of design in the form of independent views of data requirements. The analysis of existing information requirements (bubbles #2.1 and #2.2 of Figure 10) is accomplished by studying the current environment and deriving its logical equivalent. In the discussion which follows, the study turns first to the derivation and analysis of new requirements for HP users. New Requirement Analysis An integral part of the research components outlined in Figure 10 is the provision for incorporating existing but unsatisfied data needs into the design. As indicated in the previous discussion, the following three cases (which parallel Anthony's [1970] decision framework) have been identified. 1. Accounting for warehouse labor productivity. 2. Choosing the least cost distribution system. 3. The warehouse site selection decision. In each of these cases relevant data has either been missing from decision and control models or been provided by an ad hoc data gathering system outside the existing accounting structure. The three cases and the procedures used to analyze them are discussed next 0 82 Accountipgyfor Warehouse Labor Productivity Improvement in productivity has recently emerged as a vital issue to both the national economy and the individual firm.2 From the firm's point of view, advances in productivity analysis can strengthen managerial guidelines for planning, controlling and evaluating operating performance by providing insight into the sources Of change in costs, investment requirements, revenues and profits. To date, however, the traditional accounting system has been of little value in the measurement of productivity despite the fact that changes in productivity are a basic determinant of enterprise profitability. Accounting measures, such as the ratio of value added to manhours or total costs to sales, are generally useful only at highly aggregated levels. Effective management, however, also requires analysis beneath aggregate firm-wide or plant-wide relationships to the behavior of the component operations which underlie them. These less aggregate measures are 2 This is evident from the many articles in the popular press that have recently appeared on the subject. For example, Business Week, June 30, 1980 devoted a special issue to "The Reindustrialization of America" and Newsweek, Sept. 8, 1980 featured ”Productivity in America.” In addition to regular features on the subject, the Wall Street Journal's 1980—81 issue of Collegiate Forum published a special edition on ”Productivity in the United States." The accounting literature has just begun to recognize the importance Of productivity measurement. See, for example, Wait [1980] and Gold [1980] whose articles are the result of the NAA's Management Accounting Practices Subcommittee on Productivity and the May-June 1981 issue of Cost and Management whose articles centered around the issues of accounting for productivity. 83 rarely provided by accounting systems and where available are usually collected as by-products of such systems as procurement, production control, personnel and sales [Gold, 1980]. The desire to remedy this type of ad hoc data gathering led Gold [1980, p. 31] to conclude: Productivity analysis must be broadened in coverage. It also must be integrated with existing information flows and control systems, and made applicable to the task of appraising the effects of prospective innovations in inputs or processes, as well as uncovering the specific causes of past adjustments in results. Productivity and the lack of adequate data to measure it has been a central concern of distribution. A recently completed NCPDM (National Council of Physical Distribution Management) productivity study found that 85% of the companies surveyed have no program to improve distribution productivity and concluded that a principal reason for this deficiency was that "good data to form the basis of productivity measurement simply did not exist” [Stewart and Morehouse 1978, p. 14]. The study was critical of accounting systems which were found to produce predominantly aggregated measures expressed only in monetary terms. A common complaint was the lack of an adequate internal database with which to construct productivity measures. The focus of this first case is on a specific element of distribution productivity-the warehouse labor element. Choice of warehouse labor results from two factors. First, it has been 84 identified as a problem. Second, the research also intends to show how distribution financial reporting requirements can be achieved from the same data model that provides data on productivity. In recent years payroll accounting has become more complex. In addition to collecting labor information for company use, employers must now act as tax collectors for various governmental units, each of which requires numerous detailed reports. Furthermore, many employers now act as agents for employee unions by deducting and paying union dues. The combination of these factors makes the choice of warehouse labor a suitable area of study. Developing apprOpriate warehouse labor productivity measurements requires identifying the specific information needs of the users who will be employing the measure. The next section outlines these needs and is discussed in terms of the requirements of the HP Company's managers. Statement of Requirements Based on the realization that distribution costs are a steadily increasing proportion of the sales dollar, the Vice- President - Distribution of the HP Company has asked the accounting and systems departments to develop» a distribution productivity information system. Ihitial estimates regarding the net benefits of the system are promising. Current plans are to implement the system in stages beginning with the development of warehouse labor productivity measures at the PFW, RDC and LDC locations. 85 Discussions with the Vice President and individual warehouse and distribution center managers revealed the following overall requirements concerning the productivity measure. 1. 2. It must be readily understood by those who are measured by it as well as by those who use the measure as a performance standard. It should reflect the different Operating conditions existing at the various warehouse and distribution center locations. It should reflect real changes in performance as Opposed to changes which either result from movements in factor prices and inflation or result from inaccurate representations of the accounting system. It should take into consideration effects of seasonality and volume changes. It should have the capability to measure the productivity of individual employees as well as the specific responsibility center to which they are assigned. Further discussions with all levels of management were carried out to Operationalize these requirements. From these meetings, the following specific reports were requested.3 For warehouse and distribution center managers: 1. Weekly productivity reports by work area within warehouses and distribution centers giving productivity comparisons between current-week, previous-week. and this~week-last- year. Monthly productivity reports by warehouse and distribution center showing productivity comparisons between current- 'month-actual versus current-month-budget, current-month- actual versus last-month-actual and current-month-actual versus samevmonth-last-year actual. 3 Many of the ratios and reports are suggested by the NCPDM [1978] study on productivity improvement. 86 3. Weekly labor summary reports by warehouse and distribution center showing for each employee -- total hours worked this week and cumulative to date, total earnings this-week and cumulative to date, and total hours assigned to a work area within a warehouse or distribution center. For the V.P. Distribution: 1. Monthly productivity report for each warehouse and distribution. center showing actual ‘versus budgeted productivity outcomes. 2. .A monthly report contrasting for this-month, last-month, and this-month-last-year, the following ratios for each warehouse and distribution center .Compensation per man-hour .Unit labor costs For the manager of the payroll department: 1. A quarterly report for federal income tax purposes indicating total wages subject. to withholding, federal income taxes withheld, total wages subject to social security taxes, the amownt of social security taxes due from employer and employees, and, the combined income tax withheld and social security taxes due. 2. .A weekly earning record showing for each pay period for each employee: hours worked, regular and overtime earnings, total earning this week, individual deductions, net pay this week and cumulative earnings to date. 3. A weekly payroll register indicating for each employee; hours worked, regular and overtime earnings, total earnings, individual deductions, net pay and distribution of earnings. Analysis of Requirements From the various requirements specified above, the objectives of the accounting and systems group assigned to design the system can be described by the following tasks. 1. To provide various levels of distribution management with a meaningful labor productivity performance measure. 87 2. To define the data needs implied by the user requests. 3. To describe these needs in a logical format. Although the first task is not the primary focus of this research, the quality of the productivity measure is a major determinant in the acceptance and viability of the information system. If the productivity measure is constructed on a solid methodology, the reports which are generated have a higher probablilty of being acted on by managers and of being accepted by warehousing personnel. Consequently, Appendix A discusses the develOpment of a pragmatic yet theoretically sound productivity measure. Consideration is given to alternative measures of input and output, definition of work areas, and effects of volume changes. The outcome of the discussion in Appendix A is the adoption of an efficiency productivity ratio of the form, Equivalent Cases of Output Allowed Man Hours Worked and an effectiveness productivity ratio of the form, Productivity Efficiency = Actual.Cases of Outppt Allowed Man Hours Worked Productivity Effectiveness = The second and third tasks relate directly to the methodology shown in Bubble 1.2 of Figure 10. At this point, the data elements and processes required to produce the productivity and payroll reports are defined and then descirbed in logical format. Completion of the payroll requires specific information on payroll data, skills data, identification data, deductions data and performance data - mmch of which will already be found in existing 88 payroll programs. However, new data needs are evident if the reports desired by the distribution executives are to be generated. These additional data requirements consist of the following: 1. Labor cost standards which must be established for each task in each storage location (see Appendix A for computation of standardized time units). 2. Time work measurements for each employee for each task performed. 3. Output or accomplishment measurements for each employee in terms of the tasks performed. To illustrate the methodology in the following discussion, the receiving and storing work area4 will be used to describe the new procedures. Figure 12 is a context diagram which indicates the scape of activity to be studied and Figures 13 and 14 represent data flow diagrams which further partition each of the two context diagram processes. The discussion which follows outlines what the system is to do but does not, at this point, address the question of how the system should be physically configured to carry out the task. As indicated in Figure 8, issues of implementation are deferred until later stages of design. Figure 13 illustrates the specific distribution tasks that must be accounted for under the new system. Documents accompanying merchandise shipments from vendors or intracompany transfers are 4 Figure Al and Table A1 of Appendix A describe more fully the classification of work areas and job tasks. 89 $9 #7:»... 209+ con-Irol iuflzcvvuaa- . .fiuv’ can‘hwncd ' unnaJ £;M§ur ‘2. - «mnphqgg Ehnwunuu. mo LAeoe 5 Mann" unencumxn. FHuhufiivl '”“PkHF¢- ' 4h Mus” urn-f ‘ " I coo? 123" . W 'Ma*“*Wn Furor? am1::rfi,, Oflfifibna:’ f" w hour- FIPDP+ FIGURE 12 Context Diagram Of Warehouse Labor Reporting System For Receiving and Storing 9O Pku\4br kdvr coat WI hi I‘- R’ESOZD ini‘I’bh'zcd. was MM“- 4"th ARRIVAL. “mark" dally 5'5? aP' P'O‘I‘” aka! 4a k 5 Y (BMFLOYEfiéI FIGURE 13 Diagram 1 : Receiving and Storing froduedhfi+y ewe.” VAvaou. ”I“ ACCOUNTANT ”bx“ b FIGURE 14 Diagram 2: Payroll and Labor Distribution 92 matched to a copy of the original acquisition order and a receiving report is filed. Employees will record their arrival by means of a time recorder device and will be assigned specific tasks to complete. These tasks depend on the nature of the receiving function at each PFW, RDC and LDC but include receiving and storing (1) full pallets, (2) partial pallets, and (3) loose cases. An additional task is to segregate damaged cases into special holding areas and prepare them for return. Daily task recorders will be issued to each employee, the purpose of which is to record the beginning and ending times of each task performed. Also included on the recorder will be the number of cases handled by each individual during the performance of each task. At the end of each shift the time and task recorders will be assembled and a consistency check made to ensure that the number of cases indicated on the task recorders coincides with the numbers on the receiving reports. The checked task recorders are then compared with the completed time recorders to ensure time equivalence. Information regarding the daily labor transactions will then be forwarded to payroll. WOrk in the payroll department consists of computing payroll costs and distributing these costs to work areas and cost centers. These requirements are decomposed into the six procedures shown in Figure 14. Following a control process during which the information on the recorders is entered into the system and verified, two different but related procedures take place. One is 93 concerned with payroll processing and is exhibited in Bubble 2.2, 2.3 and 2.4 while the other is concerned with productivity report generation and is shown in Bubbles 2.5 and 2.6. Included here are a number of new files and extensions to existing files which are required to provide the data necessary to meet these objectives. These are now briefly discussed. Emplgyee file is an existing file whose contents are organized into such categories as skills data, benefits data, identification data and payroll data. The edit function utilizes the identification segment of the employee file, while the payroll system utilizes the payroll segment for drawing on and updating information relating to wage rates, deductions, hours worked, exemptions and other such data. Transactions detail file should contain a detailed record of the information contained on the daily task recorders. Included in this file, for example, should be data regarding employee identification, location, work area and task codes, task times, and output levels. Information from this file will subsequently be used to prepare desired reports on labor productivities and costs. Work area file will contain the data items that serve to exercise control over defined responsibility centers. (Appendix A defines these centers as receiving and storing, picking and packing and shipping). Included in this file are actual and standard input measures (time and cost) and actual and budgeted output measures 94 (cases handled). Ihi this file will be found supporting data for the cost center time and expense summaries. Task standard detail file will contain the data concerning the time required to complete a given task in a given work area in a given location (this information was developed in Appendix A). General ledjer summary file will contain a record for each general ledger account balance that is affected by the labor update program. Cost center file will contain information relating to individual warehouses and distribution centers. This concludes the description of the data and processes needed to meet the users productivity and payroll requirements. Following a similar description of the remaining two cases and the existing system, the information develOped here will be used as an input into the E-R data modeling process. Choosipgythe‘Least.Cost Distribution.System Many decisions which must be made in the distribution environment involve the tradeoff between one set of costs and another. Examples of such tradeoffs include the choice between additional stocking locations and less expensive transportation; the exchange between physical labor and the capital cost of automation; the exchange between levels of desired customer service and costs of holding inventory and the choice between shipping by common carrier or using a private fleet. In fact, the integrated distribution management concept rests on the premise that, given a 95 stated customer service level, individual distribution cost components can be integrated so as to achieve a minimum cost. As noted earlier, however, accounting systems have generally lagged behind the developments in distribution, resulting in a lack of adequate cost data required to informatively make such decisions. In their monograph Lai and LaLonde [1974, p. 219] note that: a major difficulty in the design and analysis of physical distribution systems concerns the construction of the data base. Conventional accounting and management information systems do not readily provide the types of data that are needed for these studies...In constructing a data base, the problem of non-availability of raw data is a major constraint. Christopher and Ray [1976, p. 9] state flatly' that, "Existing management accounting data does not give us the information necessary to take advantage of the tradeoff situations which exist." They further note that independent data gathering schemes to collect the information not provided by the accounting system represents "a duplication of effort and an unwarranted cost" [1976, p. 11]. Rayburn's [1981] recent survey also points out the limited role that accounting has played in analyzing distribution tradeoff potentials. Assembling a database for managerial decision making would be difficult to justify if the data involved was only required for a single application. However, many of the 'what if' type questions facing distribution executives involve common data which is shared by different users depending on the issue at hand. Disaggregated data on inventory carrying cost, warehousing cost, and 96 transporation cost is needed, for example, to address each of the following concerns. . What savings are incurred in distribution expenses if a product line is eliminated? . What is the impact on cost if shipments are made factory-customer direct rather than routing pmoducts through intermediate distribtion linkages? . What is the effect on total distribution cost if customer service levels are increased? . What is the economic consequence if the number or location of distribution centers is altered? While the data implied in these questions is oriented to questions of logistics, ix: is not distribution specificJ Issues involving product continuance, tranSportation channel choice, customer service and facility location are as important to managers in production, marketing and finance as they are to managers in distribution. Hence there exists many areas where distribution data can be viewed as shared data to be used by different executives for different decision purposes. Such 'what if' questions are also considerably more important to the executive than are routine reports on operating activity. However, due to the problems of inappropriate data and the traditional emphasis on the application program approach5 to ad hoc requests for decision information, the ability to accurately 5 Gessford [1979, p. 96-97] describes the application program (AP) approach as one which emphasizes the development of 97 assess the merits of alternative distribution systems has yet to be achieved. Events accounting (with its emphasis on multidimensional and disaggregate data collection) coupled with a semantically rich data model (the E-R model) would appear to offer a potentially significant contribution to these existing problems. To illustrate this capability, the information requirements associated with the choice of the least cost shipping mode is next analyzed. Statement of Requirements Recently, the V.P.-Distribution of the HP Company has been investigating methods of reducing the overall cost of distribution. Specific data to facilitate these inquiries has been difficult to obtain from the accounting department because 1) the data associated with most distribution activities is too highly aggregated or 2) extensive reprogramming is necessary each time a request for such information is made. A specific issue that he would like to investigate is as follows. Can a saving be achieved individual computer programs to handle routine areas such as payroll or accounts payable. In this approach, data is organized in a manner that is most efficient for the specific application at hand and is Often restricted to that application. Ad hoc requests for information have proven unsuccessful because the process of extracting the required data is complicated, time consuming (reprogramming is often necessary) and the data arrives too late to be of any value. An alternative AP approach to ad hoc requests is to provide the decision maker with hard copies of individual file contents in the hope that he can find the data elements of interest. Both approaches have proven unsatisfactory. 98 by shipping directly from PFWs to certain RDC customers via common carrier rather than by shipping via the existing distribution network -a network which first inbounds products from the PFW to the RDC by rail, then stores the: merchandise at the ‘RDC, and finally outbounds the goods to customers by common truck carrier. The V.P. has stated that customer service level is to remain constant. Thus, the question is whether the savings achieved by stocking less inventory (as will be the case if product is shipped directly from the PFW) are greater than any increases in transportation costs that may be incurred by transporting customer products directly from the PFW. In response to this possibility, the V.P.-Distribution has asked for a report which compares the yearly costs of the alternatives for a specified group of customers 0 Analysis of Requirements From the various requirements specified above, the objectives of the accounting and systems group assigned to design the system can be described by the following tasks. 1) To define the data needs implied by the V.P.'s request. 2) To describe the logical steps necessary to obtain this data and to process the data into meaningful report form. The task of specifying the data needed to compare the two alternatives is not a difficult one. In order to successfully 99 contrast the different modes of delivery, information about the following items is required. Sales - current and estimated volume of product shipped to customer. Warehouse Flow Through - variable cost of receiving and storing, picking and packing and shipping by product by location. Inventory Carrying Costs - inventory carrying cost by product by location. Transportation - inbound freight charges by product from PFW to RDCs. - outbound freight charges by product from RDCs to customer. - outbound freight charges from PFW to customer. Transportation Flow Through - inventory carrying cost by product in transit. While the specification of the data is a relatively trivial task, the job of collecting the data in such disaggregate form is far more imposing. thortunately, the literature in distribution is often silent regarding this stage of system develOpment. Because of the difficulty in obtaining disaggregated data, many models dealing with evaluation of alternatives simply use assumed data inputs to derive their outputs (Bowersox, 1974, p. 370). Such assumptions may be useful for pedagogical purposes, but decision making capability is reduced considerably if the decision maker is unable to determine the current actual cost. of {completing the distribution mission. Using assumed data clouds the fact that information regarding distribution events is often aggregated to a level where it is extremely difficult, if not impossible, to obtain 100 the type of actual cost information which sheds light on individual events. Since an analysis of alternatives presupposes a knowledge of the cost of the current situation, the prOposed system must pay close attention to the critical data collection link. In this regard, the events accounting philosophy plays an important role. Figure 15 is a high level data flow diagram illustrating the various data inputs, files and processes involved in the production of the desired distribution alternatives cost report. Processes 1, 2 and 3 provide information relating to the variable warehousing costs. Process 4 relates to the computation of unit freight cost, Process 5 relates to the carrying costs while Process 6 yields information on current and estimated customer order patterns. To further identify the nature of some of these processes, additional partitions have been developed in Figures 16 and 17. Bubble l, 2 and 3 of Figure 15 illustrate the logical steps that are involved in obtaining a ‘variable.‘warehousing cost 'by product. The procedures required to obtain this information correspond very closely to the DFD's illustrated in the previous case. In fact, the only difference is the new requirement that the labor costs attach to the product rather than simply to the work area or cost center. Therefore, much of the discussion which took place in the previous case dealing with labor productivity applies here. For example, to obtain the desired warehouse cost information, employee task and time recorders must contain information which reflects the various attributes of the events amused C-f-aaob 101 TM WW I ‘ggrmcd employee. “"39 new." 60"!» RECEIVINC? wnfi'med da; "I M p aok nccor. e,” “OQING 63%34 fifilJES 02952 FILE: saw; ea- ‘hmoo E23461; H. €¥:%ww+vr' :1. “”22 .+ dowm ' F1 (A... erabb ‘ 0209.2 W19” 13¢VENQVCW57 pxw~ca¢cL F'vl LE. 26021 P115 . by + FILE by urea-hon ‘juncuaentis» EMLL. cw: 4*€€¥"* ' FILE: ”O‘Né FtLE Invoice WOMEZ F: LB VATA report ”Hm-fee: FIGURE 15 Diagram 0: Evaluate Distribution Alternatives 102 au. 0" Mon-te- 4:: MT“ INVDIGE 1b luuc.4al- LAO! H6 9Hun~unrflb :| L; 2F| P15 Nurhdmui - 4W9 5mg“ flumflcca INVE-N‘l'a‘l 4oz. —- czauarna JUZMFF FHL ‘r I. mu “1 C057: ihu'kk 409* c* by QQéIInI/ch:3h::::;ns FIGURE 16 Diagram 4: Traffic 103 INVENTOZV Fl he. FIGURE 17 Diagram 5: Inventory Carrying Costs 104 taking place at each work area in each stocking unit. Thus, employee data relating to activities, tasks, times, finished goods and quantities of product moved are first captured on the task and time recorders. This information is subsequently processed in a manner similar to that described in the payroll and labor distribution section of the previous case-‘with the addition that costs will now be associated with specific products. Resulting from this processing will be a variable labor cost per case by product by location. In view of the close correspondence between these requirements and those derived earlier in Case I, no further partitioning of these three processes will be carried out. Figure 16 shows the partitions relating to freight information. Obtaining freight costs at the product level requires that details regarding the freight event be captured in dis- aggregate form. For each shipment received or sent, information regarding product number, unit of measure, number of cases, weight and classification codes must be Obtained. The freight invoice is initially matched with such a detailed freight document and a series of calculations are then carried out to determine a freight cost by product for both inbound and outbound shipments. The process which leads to the freight cost per product will reflect the fact that common shipment costs are a function of weight and perishability. Thus, each shipment cost is assigned to individual products based on that products contribution to the total shipments adjusted weight. 105 The process of obtaining data relating to inventory carrying cost is described in Figure 17. It should be noted that in turn each of the processes 5.1 - 5.3 in Figure 17 can be partitioned to develop the specific components underlying the inventory carrying cost. For example, 5.1 could be partitioned to show separate processes for obtaining unit variable manufacturing costs, insurance, damage and obsolescence costs.6 The inventory file is the vehicle which provides data on both the carrying cost associated with each product and the average level of product inventory on hand by stockkeeping location. The former can be extracted from the file as a specific monetary value and the latter can be obtained by averaging the quantities on hand over a sampled number of observations. A similar sampling of the shipments file will provide details of average quantities of product in transit. Process 6 and 7 of Figure 15 can be described without further partitioning. Process 6 is one where information comes partly from ‘managerial estimates and partly from. the internal system. In addition to current sales and finished goods order patterns,management, for example, ‘must specify' the decrease in total system inventory which is estimated to take place as a result of shipping from the PFW directly to the customer. Furthermore, changes in customer order frequency and size that are associated 6 Lambert (1976) provides direction on the factors which must be considered by management when developing an inventory carrying cost. 106 with the proposed delivery mode must also be estimated. This data, plus the information arriving from the earlier processes is then assembled in Process 7, and a report contrasting the costs of the alternatives is prepared. The Warehouse Location.Decision In the past, it has been easier to identify lower level information needs than to identify information needs of the strategic variety. In part, this is due to the fact that most information systems are transaction driven and consequently needs are viewed from this perspective. It is also due to the fact that information needs of higher level executives are not as clear-cut as those whose responsibilities are less broad. Furthermore, a high prOportion of many information needs at the strategic level are characterized by a non-routine nature, an infrequent use and a soft data quality. The combination of these factors have historically weighed heavily against the development of systems which support the higher level decision needs of management. Recently, however, there has emerged a substantial amount of interest and research in the area of decision support systems [Keen and Scott Morton, 1978] whose purpose is to provide higher level information to management. While it is apparent that no system will ever satisfy all or even a significant portion of the data requirements of strategic management, those strategic. decisions which are more structured may be able to benefit from the 107 development. The warehouse location choice is an example of this type of decision. It is strategic in the sense that it represents ”one of the firm's basic strategies for accessing markets and has a significant impact on such factors as revenues, costs and service levels to customers. and clients" [Buffa, 1976, p. 82]. It is structured in the sense that it is a recurring decision which can be reduced to a number of variables for which information can be provided. The problem in a warehouse location decision is to identify those variables which are viewed as being of critical importance and deciding how best to collect and incorporate information about them in an overall information system. Paralleling the develOpment described above have been new methodologies designed to focus on the key information needs of the individual manager. Rockart (1979) calls these key needs 'critical success factors' and indicates that they represent areas of activity which should receive constant and careful attention from management. The concept of the critical success factor can be utilized in determining the information needs associated with the warehouse location decision. Once defined and evaluated as to their relative importance, the data collection problems can then be analyzed. These features will now be described in the context of the HP Co. Statement of Requirements Distribution centers represent an integral part of HP's strategy for creating the most profitable link between originating 108 points of value creation and the ultimate consumer. In fact, the right location is viewed by the company's top executive as one of those critical factors which are necessary to ensure the firm's continued successful performance. In view of recent demographic shifts, changing competitive positioning, and increased business Oportunities in a number of regions, the management of HP wishes to develOp a system which, to the extent possible, can provide data to help in the selection of new warehousing sites. Specifically, HP is considering the establishment of an LDC in a particularly high growth Southwestern city which is currently being serviced by a distant RDC. The nature of the proposed change is illustrated in Figure 18. Information regarding the critical factors required to make such a decision was solicited from executives charged with location responsibility. From these interviews the following decision processes became apparent: l. The overall corporate Objective in establishing distribution sites is to choose the location(s) which minimizes total cost at a specified customer service level. 2. The choice of distribution location is a sequential decision which begins by evaluating regions, then communities within regions and finally sites within a community. 3. The variables used to evaluate both new and existing sites include numerous quantitative and qualitative information of varying degrees of importance. In view of the large number of decision variables which emerged from these discussions, a further analysis was conducted in which the variables were ranked by degree of importance. Table 4 109 PRESENT SYSTEM PROPOSED SYSTEM PFW RDC 0.00 PFW LOC Key: PFW RDC LDC Production Facility Warehouse Regional Distribution Center Local Distribution Center Customer FIGURE 18 Diagram Illustrating Existing and Proposed Site Configuration 110 provides the result of this factor ranking process while Figure 19 illustrates a Data Flow Diagram of the evaluation process. Analysis of Requirements Two important facts are made clear in Table 4. First, the majority of factors elicited are those for which no established internal source of data exists. Furthermore, given the widespread availability of private and public sources of information for many of the macro variables, it seems unlikely that the added benefit of collecting and maintaining such data would exceed the cost. Second, while numerous variables may be examined during the site selection decision process, there are certain factors which are more critical than others. Included in this category are transportation costs about which a substantial amount of information can be provided from the database. Thus one of the primary considerations of the designer is to model the data requirements associated with obtaining information about the costs of servicing both the current and proposed warehouse sites. The requirements of this case coincide considerably with the transportation data requirements outlined in Case II. In this instance however, the focus shifts from a specific subset of customers whose freight needs are satisfied by common carriage to a more complete transportation costing system which includes common carriage, U.P.S., and private carriage. Specifically, the 111 TABLE 4 Interviews Conducted to Determine the Critical Factors Involved in Location Decisions Factors Mentioned Inbound Transportation Costs Outbound Transportation Costs Labor Availability and Costs Size of Existing and Projected Market Groups Proximity to Market Groups Degree of Unionization Community Attitudes Toward Company Location Cost Taxes and Grant Structure Proximity to Transportation System Management Preference Climatic Conditions Site Conditions Cost of Living Index Building and Zoning Regulations Utilities Availability and Cost Where 1 = critically important 2 = important 3 = lesser importance Ranking wWWNNNNNNNNNI-JHHH 112 d: ' czuMdFHano 35 2%..” “so pnghuflian "'9‘” . ¢~ah-han {but and ‘NM . 3mm? Mn 4mm «hon ‘dbr' “‘45.: W - and hénr'coahr*1 *“an coats h‘ f a? 40:?“ m «\vhg W 9”“? 4.53.. . / Winks». aflihhudo “may MM...“ Ihéhda cuuiaundhqa vqu;:2L~‘nko uvHTc‘Hcfi Mae-5d.“ :39 . evaIueJ-‘oh FIGURE 19 Diagram of the Distribution Center Site Selection Process 113 accountant and designer must address the following issues in a more complete transportation costing methodology: 1. The inbound and outbound transportation costs by origin/destination. 2. The inbound and outbound transportation costs by common carrier, U.P.S., and private carrier. 3. The cost associated with a particular trip. 4. The cost associated with a delivery within the trip. With this type of information the costs of shipping via the present site (PFW-i RDC—DCustomer) can be obtained and the costs of shipping via the proposed site (PFW-+ LDC-%>Customer) can be estimated. This is illustrated in the data flow diagrams in Figure 20. Summary of New Requirements This concludes the analysis of the data and processes needed to support the new requirements. From the previous analysis, it is clear that more emphasis was placed on Case I than on the other two situations. This is not to imply a relative importance to the first case. Rather, the intent was to fully describe and document the methodological procedures for at least one of the new requirements. Repetition Of these procedures for the remaining two cases was felt to be redundant and burdensome to the reader. Consequently in Cases II and III attention was focused more on the Vat-t let—t éfifitzf"" “H9 1kANaactuaq 5:59” . “m“ W“%v’n “‘9 a Imleo MM 4~ueu *“flhfivvrhudn\ M430" . Dnuvcn dhvuro 34L; '9 M (mks. FIGURE 20 (a) Context Diagram: Transportation Costs 112uca< FILE 021V cit—e % ‘ Egg“. “P Cod? flat. 9 W'Mfi ‘H‘ cud—bum W web m+¢10ed I‘m/oi s 9 FIGURE 20 (b) Figure 1: Obtain Inbound and Outbound Transportation Costs 115 outputs of the Data Flow Diagrams than on fully specifying the attributes, data element dictionary and transform description details. The Data Flow Diagrams developed in this chapter represent the input to the E-R models in chapter 4. Before proceeding to the modeling stage, however, the existing system must also be analyzed and defined. This is the subject of the next section. Existing Requirements Analysis As indicated earlier (bubble 2.1 of Figure 9b, and bubble 2.1 Of Figure 10), an integral part of the design methodology is a description of the current information system. In this methodology, such a description is accomplished in two stages. First, the system is documented in terms of the departments through which information is passed and the procedures by which information is generated. Viewed in this manner, the user is in a position to verify how data moves through the organization. Since much of the description reflects physical considerations, this portrayal can be called a physical view of the information network. Second, the physical view undergoes a transformation which emphasizes what the system is to accomplish rather than emphasizing the methods by which it is executed. Since this orientation is independent of the physical aspects of the system, it is termed a logical view of the information network. In order to better manage 116 the detail which follows, Appendix B and C have been set up to include a complete description of the physical and logical aspects Of the HP information system. In the next two sections, both the physical and logical systems are described, but the: detail is limited to those areas where change (the new requirements) is being contemplatedi The diagrams which follow have been extracted from Appendices B and C and contain certain notations, such as connectors, which are fully described in the Appendices. A.Physical View of the Existing Distribution Information_§ystem HP Company maintains an information system which is supported by a central computer located at headquarters with connecting terminals at each manufacturing and stocking location. Information regarding orders, receipts, shipments, labor and freight are conveyed from these field units to the central computer* which consolidates the data and triggers the release of Operating documents and periodic reports. An example of how the order processing and order receiving functions are carried out and their links to the payroll and labor distribution accounting system is described next. The remaining distribution and accounting functions .are found. in .Appendix '8. Together, these diagrams constitute the existing distribution accounting information system. Figure 21 depicts the document flow and the type of 117 Mam WI “It“ “can. mu. m ma. MT“ “”6 a a FIGURE 21 Order Processing 118 information generating procedures which occur throughout specified departments during the lorder processing function. Orders originating from customers or from field units are accumulated, entered at a terminal at periodic intervals, and verified. Once verified, orders are either accepted and placed into the open order sales file or accepted and placed in the back order file pending replenishment of inventory. The Open sales order file is processed with the inventory and customer file to produce the documents which initiate the warehouse picking operations. In the HP warehouse arrangement, filling each customer order is the responsibility’ of a single ‘picker. Consequently, a picking ticket is assigned to a picker who (1) proceeds to the inventory bin locations noted on the ticket, (2) picks the indicated inventory amounts, noting any shortages on the ticket and (3) takes the orders to the packing and shipping area. Here freight documents are prepared, information concerning which is noted on the picking ticket. The goods are then prepared for shipment at which time the completed picking document is returned to the general office and accumulated with other tickets. Periodically, information regarding the shipped order is entered at the terminal, a process which serves to update the inventory and shipments file. (The shipments file is subsequently used in the invoicing printing process - see Appendix B) Other' documents associated with the shipping process include freight documents, a 119 shipment confirmation which is mailed to the customer and a packing slip which accompanies the merchandise. In a similar fashion Figure 22 illustrates the current procedures for receiving and procurement. Here, it is noted that the arrival of a finished goods shipment initiates a procedure of matching the shipping document to the original order. Once matched, employees receive and store the merchandise. A receiving report is prepared which serves to update files and initate reorder procedures. Of particular interest in these last two Figures is the treatment of labor. Chrrently, the HP Company manages its labor through the use of employee clock cards at each cost center. At present no attempt is made to relate each employee's work effort or productivity to work areas or to the specific tasks that he has been assigned to carry out within the work area. Accounting for distribution labor simply involves the collection by field unit personnel, of completed clock cards from the cost centers. Figure 23 illustrates the accounting for distribution labor in more detail. Completed clock cards are assembled at each cost center location and information. on the cards is entered at a terminal. (Input preparation and control details have been minimized for space considerations. Despite a simple terminal illustration, it should be emphasized that full control procedures for errors are in effect.) Following verification, an approved 120 diam-.1 Mm». mo MN. FIGURE 22 Order Receiving and Procurement 121 mw/ FIGURE 23 Payroll and Labor Distribution 122 payroll tape is generated which serves as the input for a computer file update. One such file is the cost center file which contains a record of the expenses associated with each warehouse location. The labor distribution which ensues provides a report which indicates the total warehouse labor cost incurred by the individual warehouse units. Following receipt of this report, plans may be made by distribution management for controlling future labor costs. These plans are transmitted to the field where they are taken into consideration during the preparation of work activities. As was noted earlier, it is this area of distribution labor reporting that the managers of HP wish to expand into a productivity reporting system. During the above discussion, particular concern was placed on where and how things take place. Departments were highlighted, automated processes were distinguished from manual processes and even distinctions regarding tape and direct access files were noted in the illustrations. Once accomplished, however, attention can turn from physical identities to a more logical view which focuses on the objectives rather than the procedures of the system. This is the subject of the next section. A Lpgical View of the'Existingypistribution Information System During this stage of design, a logical equivalent of the physical system is derived. This is accomplished by removing all references IX) physical concerns (departments, file accesses, mode 123 of processing, etc.) in order to better describe what purpose the system is to serve. Figure 24 is a logical data flow diagram corresponding to the order receiving and procurement function shown in Figure 22 and illustrates the data flows, data files, and processes which occur there. Additional detail concerning the processes of this function is provided by partitioning each bubble in the Diagram into its subprocesses.7 For example, the 'Receiving and Storing' process (Figure 25) can be partitioned into four subprocesses revealing additional details. Similarly, 'File Update and Procurement' (Figure 26) consists of five lower level processes. It should be noted that in every case the diagrams are balanced in the sense that the net data inputs and outputs of the higher level partitions exactly equal those of the lower level partitions. Figure 27 is a logical data flow diagram of the Payroll and Labor Distribution function. Again, it can be seen that although the information corresponds precisely to the physical analogue the focus is on what the system produces rather than on the mechanics behind its production. 7 The numbering scheme ‘used in these figures provide a systematic way of proceeding from higher order abstraction to lower level detail. This is made clear in Appendix C. 124 Ffiwohcd IEuawa «MW W W- "3 * a? u 11...?" lfivhérIFann ‘br'hflbor cosi- candnal Ehcefiinmg Eipvrfl' Auuauuanncnq owes FHA W27 :‘Lt flzlcauanes F' , .fiAanufhchufitg ' V‘aHCIIZ Faruivwse. I Cindawab Qphwit'wmf Odor-9 FIGURE 24 Order Receiving and Procurement Note: Numbers in Bubbles Refer to Numbering Scheme DevelOped in Appendix C 125 daupuuawuanu 02m File: lafifi$+uhfl Ghani’ ’fizum’fier 1 ldkun-Cb§:' gzawdcw- Cbnfiv" PFW?” Chum Mal-dual an.» Enkzluvnreea T! .'a ha...) |2¢C5$¥E J9“? SWDIHE mm” ICGnJEnnnceI EmPk’Y“ ~ l fine» ““5 “a W fibfor+ FIGURE 25 Receiving and Storing Note: Numbers in Bubbles Refer to Numbering Scheme DevelOped in Appendix C 126 anzeri' OOEN mnenfim CEH>EIZ FHLE. ZECEJOT'E FWLJE <29u61¥5r517 IerZnJL «JFEaKrE troverrrofih/ AND WI-NT amzxzcmaz LMST Inqveuqrenzv' I 250206.12. 6:! LE “an: AMNAe-EMENT Qaquia'ahbne g§::::;c: ‘rgtghoe. 6112 :56 @9234» s‘W .. , L. 2%:%2::::' Adarureuaf ' ‘ W n4 om n 23:52::: Clwdcwws ‘9 FIGURE 26. File Update and Procurement Note: 127 6W F‘LB pAYROL-L Fl LE WI ‘I-D 'JN AAIUM115 202. + REZflDGU’ IHLJE 6245:345» HHIH‘ E5 as e m 7“: $95 ”9 VAYBQLL. ACCOUNTANT FIGURE 27 Payroll and Labor Distribution Numbers in Bubbles Refer to Numbering Scheme Developed in Appendix C. 128 Conclusion This chapter has laid the foundation for applying the entity- relationship accounting model to an organization which has many managerial as well as financial accounting data requirements. A research orientation to the Lum e£.gl. (1979) and De Marco (1979) methodologies was described which allows (l) a hypothetical case example to be used to describe existing information needs and (2) a review of the literature to substitute for actual interviews in ascertaining new data requirements. Three examples where distribution data needs have traditionally not been met by the accounting system were then drawn from the literature. In each case, a new system to alleviate the data deficiencies was proposed, and a logical model of the system's objectives was described. The existing system was then illustrated and discussed. First, the analysis focused on identifying the physical characteristics of the information flow. Second, a logical view was derived from this physical representation resulting in a portrayal which was unencumbered by physical identifiers. The specification of the existing and new information requirements leads naturally to the next step of the design procedures shown in bubble 2.3 of Figure 10. Here the methodology combines the logically derived existing needs and the logical assumptions of the new requirements into a comphehensive model. This is accomplished with the use of the entity-relationship accounting model and is the subject of Chapter Four. APPENDIX A 129 APPENDIX A DEVELOPMENT OF THE PRODUCTIVITY MEASURE It is evident from the requirements outlined by the distribution executives that, prior to any systems changes, careful consideration must be given to the choice of the productivity measure. Since productivity is a term which has been used in many different contexts,1 it is important to define the use of the term in this study. The most general form of the productivity ratio is given as equation [1]. Real Output Real Input Productivity - [1] This study will concentrate on a particular form of the ratio, namely, Real Output [2] Man Hours Justification for this form of the productivity ratio is based on Productivity - the following criteria. 1. It is in demand. 2. It is useful in monitoring current labor performance and in estimating future labor requirements. 3. It is related to wage and labor cost. 4. It is easily understood by those who monitor it or are monitored by it. 1 Various types of ratios often computed from accounting data collections are called ”productivity ratios". For example, labor expense to sales dollars or operating expense to sales dollars are included in this category. Each of these, however, incorporates into the inputs and outputs, none constant factor prices and changing inflationary conditions which reduce the reliability of the ratio. 130 The denominator of equation [2] may be measured in several ways. For example, man hours may be viewed as equivalent to "man hours paid" and as such can be readily obtained from the accounting records. Greenberg [1973] and Bucklin [1978], however, both criticize this measure because of the uneven effect that changes in work and leave practices have upon its calculation. To illustrate, if a work. period is increased. or decreased due to changes in workloads, the "man hours paid" also rise and fall accordingly. This is correct. However, if the work period is decreased due to changes in vacation and paid leave policy the "man hours paid" measure does not change. Consequently, a decreased workload and an increased vacation leave both effectively reduce man hours worked, but are accounted for in a different manner. Further, the difference between the two is an accounting difference, not a real difference. A second and more consistent measure of the denominator is "man hours worked". This measure covers all hours on the job including coffee breaks, fatigue and down time. It excludes all leave time whether paid or unpaid. While such a measure overcomes the problems of the "man hours paid" statistic, it is more difficult to obtain, particularly for salaried employees. Nevertheless, its conceptual superiority* argues for using this interpretation of the denominator. The numerator of the measure must be versatile enough to reflect the output of different situations. This requires 131 attention to three tasks. First, the specific work areas for which a measurement is to be made must be defined. Second, a measure which accurately portrays the output of the defined work area must ‘be developedi Third, the additional data requirements associated with the measure must be collected and integrated into the information system. The first two concerns are discussed next and relate specifically to the HP company. Definition of Work Areas In virtually' any ‘warehousing situation, certain tasks are routinely performed. Evaluating the performance of the individuals working on these tasks dictates that work areas or responsibility centers be established and that data about individuals be collected by such work areas or responsibility centers. How the areas are defined will vary with the nature and size of each organization's warehousing facilities. A large multiproduct distribution center leads to many employees performing homogenous sets of tasks resulting in numerous specific work areas. On the other hand, the management of a smaller warehouse may require their employees to perform more heterogeneous tasks resulting in a broader definition of work areas. Despite these variations, however, there will be certain logical groupings of activities which may be accounted for as a separate work unit. In the case of the HP company, the same sets of activities regularly occur at each PFW, RDC and LDC. Diagramatically, these 132 are shown in Figure Al beginning with the receipt of goods and continuing until the goods are shipped to either RDC's, LDC's or customers. The diagram also illustrates that those activities which require close coordination can be clustered and treated as work areas. This exists, for example, in the case of receiving and storing and in the case of packing and shipment. The picking activity, while not independent of other activities, is unique enough to warrant a separate accounting. Definition of Outmut in Each Work Area Although the basic nature of the activities taking place at each PFW, RDC, and LDC are the same, there are some important differences in the manner in which they are carried out. At the PFW cost centers, for example, inventory is received in single cases via a conveyer belt linking the PF's and PFW's. The cases are then palletized and moved by mechanized vehicle to predetermined storage slots. At RDC and LDC cost centers, arriving inventory consists predominately of full pallet loads with smaller amounts arriving in partial pallets or as single cases. On arrival, the goods are either stored directly into predetermined locations or are reassembled before they are stored. Similarly, the packing and shipping operations of the RDC's and LDC's differ from those of the PFW's. At PFW locations, a high proportion of shipments are palletized full loads requiring less labor effort per unit shipped than is the case at RDC and LDC locations. At these 133 mCOHumOOA kHOuco>cH um mmuauoooum mo mocoavom wcfiumuumSHHH Emuwmfia Hoaoumao xmme ammo vomemQ xmmH ammo omooq xmme Omaamm Hmauumm xmme umaamm Hash mxmwe wcfimmasm can wafixomm wcaxuam wcHHOum mam waw>fiooom mmou< xuoz mmfiuaafiomm coauoawoum new Hoodoo coHuanfiuumHQ Hmooq Iona voodoo SOHuooHu Iowan HMCOmem loam omaoeoumz xufiafiomm coauoanoum lame muoucoo umoo muouswo mwaawnfimaomwom nomco> H< MMDUHm one egg one O" AIII. mmooo mo 30am mafia cm want wcaxuam . _ wcfiuoum mam waH>Hmomm mou< xuo mm 134 latter locations, shipments regularly consist of partial pallets or single cases and require different handling techniques. The determination of an acceptable level of performance in such diverse conditions is a difficult task. NOt only do the work practices at various cost centers differ but the mix: of work practices may vary within a single cost center. In addition, changes in work; methods occasionally' are made in the hope of improving labor productivity. Consequently, any measure of output adopted must take these variations and changes into account. This requires a distinction between productivity gains and losses due to effectiveness2 and productivity gains and losses due to efficiency. In locations where work methods are reasonably stable, the overriding purpose of the accounting system is to monitor the efficiency with which individual activities are carried out. (h: the other hand, if structural changes are considered or adopted, the accounting system should reflect the productivity effectiveness associated with the change. At issue are the following two questions: 1. Given an accepted set of work methods, what should the productivity level be? (efficiency) 2 Effectiveness is defined as the ability to reach a desired objective. This objective may be to improve labor productivity through improved work methods. Efficiency is defined as the optimum relationship between inputs and outputs, given a specific set of work methods. 135 2. Given a change in work methods, what difference should occur in productivity? (effectiveness) The first question deals with control and seeks to measure the efficiency with which individual activities are conducted. In a warehousing environment where the same outputs require different labor inputs (picking lOOO palletized cases by forklift requires less labor input than picking 1000 loose cases by manual picking methods) the efficiency measure must recognize the existence of these diversities. In other words, individual managers must not be penalized for the work methods currently used, provided their efforts are efficiently carried out. The second question recognizes that there may be a better (i.e. more productive) way to produce the output, even though the current operation is performing efficiently. For example, the ability to periodically ship full pallets of inventory to customers (rather than shipping a case or two at irregular intervals) would lead to higher productivity even though the current procedure is being carried out efficiently. Developing a useful measure for the efficiency question is the more troublesome but can be accomplished by drawing from the literature on process costing. Process coating is a type of cost system found in a homogeneous manufacturing situation. A key feature of this system is the measurement of the outputs of 136 particular process or work area in terms of the number of equivalent units of production completed. Adapting a variation Of this feature in the warehousing environment can lead to a more consistent measure of output which in turn leads to a more useful productivity measure for control purposes. Operationalizing the concept first requires that an industrial engineering study be conducted to determine the time necessary to complete certain tasks. On the assumption that such a study has been conducted at the various inventory locations of HP Co., the following will highlight the develOpment of an equivalent measure of output. The example will also serve to illustrate how the effectiveness question is answered. An Example Time motion studies were conducted at representative PFW's, RDC's and LDC's to determine the time required to complete various tasks in each of the following work areas: (1) receiving and storing, (2) picking and (3) packing and shipping. The study produced the statistics shown in Table A1. Specifically, the Table first illustrates the units of time required to handle a case of inventory under different circumstances. Next, the time required to handle a single case on a full pallet in receiving and storing is standardized (set to unity) and the times to perform all other activities are expressed as a factor of this standardized value. For example, it takes an average of 8 times as long to 137 TABLE A1 Table illustrating the results of an industrial engineering study to determine the time required to complete various tasks in the receiving and storing, picking, packing and shipping work areas of representative PFW's, RDC's and LDC's. Receivingyand Storing Full Partial Loose Damages & Pallets Pallets Cases Returns Unit of Time Required to .2 .4 .625 1 Handle 1 Case Standardized Time Unit* 1 2 3.125 5 Picking Full Pallets Loose Cases Unit of Time Required to .25 1.6 Handle 1 Case Standardized Time Unit* 1.25 8 Packing_and Shippipg Full Pallet Partial Pallets Loose Cases Units of Time Required to .2 .25 .8 Handle 1 Case Standardized Time Unit* 1 1.25 4 *Standardized Time Unit Chart is derived by the following relationship: Unit Time Required to Handle 1 Case UnifieTi e Required to Handle 1 Case on Full Pallet n ce ving 138 manually pick a case of inventory as it does to receive a case on a full pallet. Expressed another way, 8000 palletized cases could be received and stored in the time it takes to manually pick lOOO loose cases. Table A2 reports the actual operating results for a period of interest for a representative PFW and RDC. Using the standardized time unit from Table Al, the actual output in cases is converted to an output measure expressed in equivalent cases. A key feature of this methodology is that it uses the same productivity performance standard for each cost center; for each work area within a cost center; and for each task within a work area. For illustrative purposes this is shown in Table A2 to be 5 equivalent cases per unit of time allowed. Contrasting the units of time allowed to perform particular tasks with the actual time required to complete such tasks, allows an actual productivity measure to be compared with the standard. The actual productivities can be aggregated across tasks to give a productivity by work area; work areas can be aggregated to give productivities by cost center; and cost centers can be similarly aggregated to result in a firm wide measure of efficiency. Furthermore, the loss due to inefficient Operations can be approximated in each case by equation [3]. For example, if the actual Actual Loss (Gain) Due 8 _ Productivity (Actual Compensation)[3] to Inefficiency Standard Paid Productivity 1139 m.N mn.~ Ansav accccauuuumeu zua>usu21ouk Anew» mo uucammawauw unoco>wuuuuun m~.o no¢.c oo~.q «cq.e Ao.nv socuauauuu huu>nuusvoum assau< an.nv socoauauum Aun>wuu=vcum nonhuman Aos«u no u«c=\muuuuv Nocowuwuum onn.N oon.~ oom.- oo~.a om oo~ 00m can om~.H oo~.~ q. ooo.o. coo.n com.” oo~.~ N. ooo.o ooo.0 oeo.e cam.o onn.~n coo.os ch 00 can co ooo.~ on~.e nae. on~.~n n-.n coo.o~ HeuOH nausea“ a mowaema oommu among muommam "nausea summaom “use "much census: a nowaaao coauu mecca auoHHam suunfiam dungeon Hugh Auvuouou usausaouua souuv vouusv no: menu no cums: Husuu< as a He uuao~H< «any to mass: “a vague saucy oaeu use care we some: aw x «V sumac ucOHa>uavm “a oHaae auras sank coauvuwvcaum Aaououou wewusaouoo sonny uo>quuom among .02 .0 .n .e .n .N saws xmnh wcquum can new>wwuou season couuaouuuauo Hasemwox cam owaoeouaz Auwuuuum cauuuavcum 3mm aou< sac; scuuaooa mmhzmu zouhamumhmuc A<20HUNM < 92¢ mmno=mm<3 rhuAHu~h<fi2mmmmmmm < mom mmmamHHU:ooxm OZHMOHm 92¢ ozH>Huumm ozHH<¢Hm=AAn m4n<fi N mam unwamu: new!» {Ea Ema Fun . emu: mu? A“ “W b“"”"”" AN04~ane»——:> annexes K_,,z’”‘ emanates FIGURE B4 Accounts Receivable 151 meet / mum W940 WM swam Me.“- I FIGURE B5 Payroll and Labor Distribution om gUHm SE .. in! ~ Fl. ,. NIT Egg @Q A NI II éo is §> g 3 §§ $7. .Fzg . gig H ZoFsoNhsL «Dramas 153 .ACCCZJBTVUQGB FIKJC£325HQ£9 HNAC TRIAL. Mo CERJTEHZ FHLJ! (ZZDT CEHQTEHZ Fiflrif FIGURE B7 General Ledger APPENDIX C 154 APPENDIX C Existing Information System - A Logical View The diagrams illustrated in this Appendix represent logical derivatives of those in Appendix B. Each data flow diagram is a 'parent' or 'child' of another diagram and in many instances can be both. The highest level of abstraction is shown by the Context Diagram which represents the scOpe of the system under study. The diagram is subsequently partitioned in a systematic manner until the desired level of detail regarding individual processes is reached. An effort has been made not to overburden the reader with detail; therefore, some of the processes have not been broken down into their lowest level components. Table Cl has been prepared to aid in the understanding of the detail which follows. 156 \aeru>amz and , fihné+fanc I7E’TEUEMJ1TCNQ AKNZCMJFJ1WfJGB ‘2F, CUMIHZAEIL FIGURE Cl Context Diagram: E3" of Ladies c.3339»- %’a LL96. fiécand RECEflFW15 FHLE “HOME—2. Cum Eénfi+tanc¢9 157 Ream 8.119 LJJZti Infivoh3¢a 6h; 3 I??? Q? | FheJ Iwufice’ Farcbwuhe Manufi‘JLl-V'lhfl Rn‘lskcd DbTQJBuTION $0M: - . |r¢FZZZAAfiGWCDLJ nwnsfierfaqa FEIXZESXSHQ69 flauPAAEAnE FhLE VGHAG V thi 4+ «nee» eve vhf-am e, Vandal“ Jimkaee Employee Tau [Esp-4‘5 @e "M . anixm Figure C2 Diagram 0: Distribution Accounting 158 EL“ 9 Ride IJMHSB Cflfigfa . . (3‘5hcwnenr LLFH5 | ;‘ fgvbw_ CznflWMthdWI . Ccrfiirhmed J Gimp cc» CUEflUMAEIL p| ‘fiww: fider: r fihqgana Wow gflJE F%M:Ffd ‘ {Nab N W ‘9 2 O ’ CEEEKL R££vauw3 - Juan I pflfierflg VEQQDENZ EEZHQFWES Fhuék+ 47» $90.45 faier9+ CbndVaI . Labor Rm” H+ Cbb+ Bill3 Qty-+5 ”Primal Qfl‘br-mance, Fhmd Fhea3h+ Echr+9 Ikuowub Ilnmmcesv 11.2 6. gm ° .me. 13w¢m¢¢v E / FIGURE C3 Diagram 1: Distribution Information Processing 159 ‘iphw;55nuw1§ (firshohmsr' ' ad“ |.|.| W Entree: C‘flPlflz CUfiflDMAIIL ave»: an” F“!- CIZCIHZ.FHLJL EHLUflIGl‘LJCIFES nus , 2v than» saun~uuai ON Hue __ Hue hI-fi. .23. ”33'; PP' €3q> . "3;" an a? .. ““3 “m: IJJis. ’Fhuflt ‘a~‘”1‘ JthufharI fflfipmnenfl' Len-af’flii «and; FIGURE C4 Diagram 1.1: Order Processing 160 Rllabb 04mg: FHLE. H liable flack OPEN ma- CIEGEHZ. FHLAE FIGURE C5 Diagram 1.1.1 : Enter Order 161 m M . 0293: race wafers: Law .04” FIGURE C6 Diagram 1.1.2 : Fill Order leblned 6&2249» Tnyufiér- nmsrft Véyuior- fi*upqfi¥w3 ticouncnfi‘ ACQQIDH’ION cazDEflZ FHLE; .I .1 AAaruf633hqu}!5 Osaka-5F Fauxflhaamt Gale—r5 @Peaekmen—r Order-5 162 FIGURE C7 Diagram 1.2 : ‘R Em Pics/cc. TInnea 5’ Order Receiving and Procurement 163 RBIs-Red Irecwla mafia-x 4a?” Cbcumeni maet'zmglset: Han Gav- WW3. - cl???” (zthVWDI WW Odors a? EFL, asses/me newer FILE l.:z.g.4, FGZIXflD IBMPUDYEE- perms ', Einpfloqce. Eéxxh~flrp5 ,'E¥“‘:1dc*1; FIGURE C8 Diagram 1.2.1 : Receiving and Storing 164 Faqflbwfiiediwuuwf' FIGURE C9 Diagram 1.2.2 : File Update and Procurement FIGURE C10 Diagram 1.3 : Traffic 166 we» mm, 132 cumez tag “are” 17"" fiWflFhAEEJT FELEB Inmate 2-l Geek: Clefltwngr Giwanwnani one: - W15 MW EEEHBVIEEI. . Cuehaneu- Confirnwui a? he“ Eéfarf Einsndcawb szsr CIH¢T35L FHLE. Rm Ihwficeo W FIGURE C11 Diagram 2: Accounting Information System 167 6H|PM¢NT§ FILE 2-I-l PRINT varuu tau rza'mt. Bite. 2-1'3 m W was Heme] FILE. Fibs y MI Vouchers owl-mo: mthg 609% Fit—E. accounrm‘t FIGURE C12 Diagram 2.1 : Accounts Receivable 168 FIGURE Cl3 Diagram 2.2 : Payroll and Labor Distribution 169 m ACQJIOIT‘ON C‘flDEHZ. F3L4£ JD~+rauapnaFaun7r Jimuom:¢es Véwmdov- Invoices IZEKZEHI’FED Hue. Faiflfil> A~aaesr FHLJL venom.— FELIL éflbifliUN. Llflafiflfll W Flt—a AttflfiDflflZJHD Fafiflfifliflflfl15 a. F‘hFVT Chaecues :Exuwuml thclwur ‘annethb lebiu—r ”W9 . WW Came: QmI-H-ance. WANT FIGURE C14 Diagram 2.3 : Accounts Payable 170 GENERAL-m2. MPH—£5 MI Voucher 2.4-a _I¢exzr DI e/ V 1 1, @ml Gianerzml Ledge—r FIGURE C15 Diagram 2.4 : General Ledger 171 References Anthony, R. (1970), Management .Accounting» Principles, revised edition, (Richard D. Irwin, 1970). Bowersox, D. (1973), Dynamic Simulation of Physical Distribution Systems, (East Lansing, Mich.: Division of Research, Michigan State University, 1973). Bowersox, D. (1974), Logistical Management, (MacMillan Publishing Co., 1974). Bubenko, J. (1977), "The Temporal Dimensions in Information Modeling," in G.M. Nijssen, ed., Architecture and IModels in Data Base Management Systems, (North Holland Publishing Company, 1977), pp. 93-118. Buffs, E. (1976), Operations Maaagement: The Management of Productive Systems, (Wiley, 1976). Christopher, M. and D. Ray (1976), Costing, in Distribution: Problems and Procedures, (MCB Books, 1976). Cushing, B. (1982), Accountiag Information Systems and Business Organizations, 3rd ed., (Addison-Wesley, 1982). DeMarco, T. (1979), Structured Analysis and System Specification, (Prentice-Hall, 1979). Gane, C. and T. Sarson (1979), Structured Systems.Analysis: Tools and Techniques, (Prentice-Hall, 1979). Gessford, J. (1980), Modern Information Systems, (Addison‘Wesley, 1980). Gold, Ba (1979), Productivity, Technolqu and Capital: Economic Analysis, Managerial Strategies and Government Policies, (Heath-Lexington, 1979). Gold, B. (1980), "Practical Productivity Analysis for Management Accountants," Management Accountipg, (May 1980) pp. 31-44. Keen, P. and M. Scott MOrton (1978), Decision Support Systems: .An Organizational Perepective, (Addison-Wesley, 1978). Lai, A. and B. LaLonde (1974), An Analysis and Evaluation of Data Base Requirements for Physical Distribution Studies," Physical Distribution Management, (Volume 4, No. 4, 1974) pp. 217-248. 172 Lambert, D. (1976), The Development of An Inventory Carryig Costitfi Methodology: A Study of the Costs Associated with Holding Inventory, (Chicago, National Council of Physical Distribution Management, 1976). Lum, V., S. Ghosh, M. Schkolnick, D. Jefferson, S. Su, J. Fry, T. Teory, and B. Yao (1979), "1978 New Orleans Data Base Design WOrkshop Report," Research Report RJ2554 (IBM Research Laboratories, San Jose, CA., July 1979). National Council of Physical Distribution Management [1978], Measuriag Productivity in Physical Distribution - A $40 glion Dollar Goldmine, (Chicago: National Council of Physical Distribution Management, 1978). Rayburn, G. (1981), Marketing Costs - Accountants to the Rescue, (Management Accounting, January 1981) pp. 32-42. Rockart, J. (1979)," Chief Executives Define their own Data Needs," Harvard business Review, (March-April 1979) pp. 81-93. Spaakman, G. (1981), "The Management Accountant and Productivity anrovement: Responsibilities and Techniques," Cost and Maggement, (May-June 1981) pp. 2-8. Stewart, W. and J. Morehouse (1978), "Improving Productivity in Physical Distribution: A $40 Billion Goldmine," in Proceedings of the Annual NCPDM Conference, (Chicago, National Council of Physical Distribution Management, 1978) pp. 1-33. Teory, T. and J. Fry (1980), "The Logical Record Access Approach to Database Design," Computing Surveys (June 1980), pp. 179-211. Wait, D. (1980), "Productivity Measurement: A Management Accounting Challenge," (Management Accounting, May 1980), pp. 24-300 CHAPTER IV MODELING THE DISTRIBUTION ACCOUNTING SYSTEM The previous chapter has provided details about the information which is currently being used by the HP Co. and details about the information which is desired but which the accounting system has yet to produce. Identifying such present and future needs constitutes the requirements analysis phase of design that was portrayed in Figure 10 of Chapter 3. As illustrated in that diagram, this chapter will use the information from requirements analysis to develop an accounting data model capable of supporting both the existing and new information demands. Chen's entity-relationship model and DeMarco's data description and transformation methodologies will be used to develop the distribution accounting model. To describe these procedures, the chapter is structured in the following 'manner. First, the information from requirements analysis is used to identify and model the entity and relationship sets for each of the three cases and for the existing system described in Chapter 3. The attributes associated with each of the defined entities and relationships are specified at this point as are the required data dictionary definitions and transform descriptions. Second, the individual views of the data are consolidated into a global E-R model resulting in an enterprise view of the conceptual distribution accounting scheme. During this stage 173 174 overlaps and inconsistencies between individual applications are noted and modifications to the data transformations and dictionary entries are made. This two part Information Analysis and Definition process, is illustrated in Figure 28 and represents a partitioning of Bubble 2.3 of Figure 10 in Chapter 3. View Modeling and Modification During the next several sections, the design needs associated with each of the three test cases and the existing system will be specified. It may be appropriate at this point to introduce the symbols which will be used during the subsequent modeling procedures. Several of these were described in Chapter II; others represent refinements to the E-R model. In Table 5, entities are represented by rectangles, relationships by Idiamonds [Chen, 1976] and attributes by a single line culminating in a small circle. Key attributes (to be described shortly) are indicated by a darkened circle [Atzeni et_al3, 1982]. In certain cases, entities can be generalized to a superentity. This generalization procedure is shown in the table as a "magic carpet" [McCarthy, 1982] rising out of the paper into 3 dimensional space. Each of the entities within the carpet may be viewed as a subset of a more general entity. However, users will often require information about the specific 175 TABLE 5 Description of E-R Symbols Used in the Study Entity Relationship * Key Attribute 14) Non-Key Attribute Generalization 176 .XL)o£nJ+a:1 anidil .0n+kPhxto rxahuoewfifimfl ANIOJumu 28mg... 3433. _ax¢mv_ «:0: OH shaman so m.N manage me :ofiuaunme a ma swnmmfia mane swflmon mo ommnm coauficawon cam mfimzamn< coaumsuowaH one mm MMDon 177 attributes of individual entities. Both views are accommodated in the generalization schema. It is also appropriate to mention that since a substantial amount of detail is involved in each of the cases under study, the chapter focuses primarily on the new and existing needs of one of the cases - the warehouse labor and productivity system. By concentrating on one of the situations, the individual components of Bubble 2.3 can be fully illustrated and the chapter can subsequently concentrate on the integration of all the cases rather than specifying similar details for each situation. View Modeling and Modification-Case I In this section, the new and existing needs of the warehouse labor productivity system are analyzed and modeled using the Chen and DeMarco methodologies. The procedure starts with the completed data flow diagrams that were produced for each application in Chapter 3. Not only do these diagrams provide explicit information about data flows and data transformations, but they also help identify the nature of the entities, relationships and attributes of the specific object system to be modeled. Resulting from this process will be the five sets of outputs shown leaving Bubble 2.3.1 of Figure 28. First, a model of the data required by each process shown in the earlier DFD's (Figure 13 and 14) will be developed. Each of these processes represents a local view which will be modeled in terms of its entities, relationships and attributes. 178 Second, dictionary entries of the data elements (corresponding to the attributes in the E-R diagram) which make up the data flows and data files of each process in the DFD's will be made. Third, data transformations describing the pmocesses which change input data into output data will be illustrated. Fourth, a new logical data flow diagram incorporating the existing and new requirements will be produced. Fifth, but not described here, are a list of constraints imposed by the users on the process. For reader convenience Figures 13 and 14 are reproduced as Figures 29 and 30. Data Modeled Views Process 1.1 - Match Receipt Document to Acquisition Order For each shipment of finished goods inventory that is received, the HP company needs to verify that (1) an acquisition order was initially issued for the goods received and that (2) the ordered finished goods inventory' is the same as that shown on either the vendor shipping document or the finished goods transfer document. An E-R diagram of the data needed for this process is shown in Figure 31. Explicit in this diagram are these entities and relationships. 1. FINISHED GOODS INVENTORY, FINISHED GOODS PURCHASE ORDER, FINISHED GOODS SUPPLIER SHIPMENT, SUPPLIER, PURCHASING and RECEIVING AGENT represent those entities about which data is to be recorded in the database. In this view the 179 Acqwmvnc»: 02052 FILE. MWISTION 0:20:12 PQEHARE QECEAHNG- Hanger FIGURE 29 Diagram 1 115459 MTH VNL‘I ‘TAbiL 003* WI l-i KW ini‘HaIs'tzd Wee Mable—€- *Mc mans/AL. W‘s" daik’ liar d" PW‘” ch61 4a k ? PDDYEEé Y I EMA Receiving and Storing 180 of" mod Emulate” confwmmulctfikl 1% f I V523?! CATION AND venuu.lflte In»: vufifnkb dad 4'39 WARE: PEnFoewuwuce ‘55, tufi+ kink- coo+ I . zaumza rue hhnu' ounnl coat « ., 1'2 °"W*“F”' 'mwdtr w... W than how chef—lab Fnbi. 'V' nnuwfl' t§;:§:““1 venfi+hu~ub ”'3' 'I ‘1’ fhaduedhfiiy .19 efiflxarusnaea FfiwwunJ. W ACCOuNTAN‘T ”Vii?" 9» FIGURE 30 Diagram 2: Payroll and Labor Distribution “0‘ FtNIéHEp Clxfis caJca:kuxfiua 0205.2 Oder In Line No. 0 @0434?” OF Gus-09 0 n F; . 0- i éaM: FflfléHED’ 1“”qu 8b @0005 INVENTozy 5M. Lino Ha, Wank] 0" Cam $910.?er , 9"- FlNléHED idea? I ' NO. I‘fCO-fhan E : . suppuez. 3M 0" 9Ht?MENT’ Co 181 Cab? C‘cn’fer 9“ Code, No. Coo? Can-her Name. 0—- PFW Mire” Vendor Na. Vanda N r “ma VENDOR Veafldor‘ Mfg/fl {:u lia- “9 HIP 3,39?le ...ofiuf-plwlr Name. --Oé.t Ref (1.55 FUZCHAfiING‘ ‘ Aeswr {21' , iNamc. 590063 Audi'hofl IO. N 5‘3""59 Au/Ivhorify 1.9. f ? mm EECEJVINC-p WT h Inn-I'Y 4'0 I FIGURE 31 E-R Diagram of Information Required By Process 1.1 " Match Receipt Document to Acquisition Order " 182 entities PFW and VENDOR can be viewed as categories of the higher level and more general entity SUPPLIER. (a) The ORDER LINE ITEM relationship describes the location on the order and the quantity of each finished good ordered. Shown as an m:n mapping, the relationship indicates that a finished good may appear on many purchase orders and a purchase order ‘may contain many finished goods. (b) The SHIPMENT-LINE ITEM relationship is of m:n order and indicates the location on the shipping order and the quantity of finished goods inventory shipped. (c) The RESULTS FROM relationship is of 1:n order to indicate purchase may involve more than one shipment (as would happen if a portion of the order was not in stock) but that a shipment is defined to be a single purchase order. (d) The PARTY TO relationships link the internal and external participants of each transaction to a corresponding event. For example, a finished goods purchase order involves both the purchasing agent (internal agent) which initiated the order and the supplier (external agent) whidh filled the cuder. The relationships here are of nzl order indicating that a 183 specific finished goods purchase order may be initiated by only one purchasing agent and sent to only one supplier but a given purchasing agent and a given supplier may be party to many different finished goods purchase orders. Each of the entities and relationships shown in Figure 31 contain properties or attributes which serve to characterize each set. To uniquely identify an entity or relationship set requires choosing one attribute from this set of attributes and designating it as the key identifier. To qualify as a key, the attribute must illustrate the following two phenomena. (McCarthy, 1979, p. 678). 1. The key must be universal - every entity within the entity type must have it as an attribute. 2. The key must be unique - no two entities may exhibit the same key attribute values. In the figure key attributes have been designated as lines ending with darkened circles and non key attributes as lines ending with clear circles [Atzeni 32.313’ 1982]. Attributes, other than those shown, could be attached to the entity and relationship sets. However, those indicated in Figure 31 are the ones which are of importance to the specific application at hand. Thus, while other applications may require a knowledge of a particular finished goods cost, color or reorder point, these attributes are not a necessary ingredient at this point in the analysis. 184 Furthermore, while each entity set consists of one or more attributes, several of the relationship sets contain no attributes. In such cases, the purpose of the relationship is simply to describe the correlation between two entity sets, and its key is derived by concatenating the primary keys of its adjoining entity sets. Other relationship sets contain additional information which is of value. It is here that information about attributes functionally dependent on the simultaneous occurrence of two entity sets is recorded. iFor example, the relationship sets order line item and shipment line item, respectively contain information about the quantity of finished goods inventory ordered and shipped. The previous discussion has described the procedures for obtaining an E-R diagram from the results of requirements analysis. In the subsequent view models, the same procedure applies. Consequently attention will be focused on the derivation of the E-R diagrams themselves and only brief explanations designed to clarify certain issues will be made. After modeling each local view a summary of the identified entities, relationships and attributes will be made. Process 1.2 - Record Employee Arrival Figure 32 illustrates that an event, warehouse employee service, is related to both a specific warehouse employee and cost center in an n:1 fashion. A particular unit of employee service is characterized by the interaction of only one employee and the cost 185 if," "m M 7 f fNO. ubuzerk24£xa . WEE lhwwVal I1?” I u0u22}uxtbe. WEE. " fiaszza. CO” Can-5e" 6“” NO“ Cos» Cervica- . I it... C0351" CENT ER FIGURE 32 E-R Diagram of Information Required By Process 1.2 "Prepare Work Activities" 186 center in which he works. However, both the employee and the cost center are party to many different employee service units. Process 1.3 - Assign Employees to Task During this process, employees in each cost center are assigned to specific distribution tasks within a work area. In the HP set-up a distribution task is accounted for at the individual employee level. That is, while many employees may work on distribution tasks, an employee is accountable for his performance on a given distribution task. Consequently the relationship between distribution task and employee indicates that a particular task is performed by one employee. As indicated in Figure 33, these tasks are of four different types, each type requiring a varying amount of time to complete. Process 1.4 - Mark Time That Task Begins The model of the data required in Figure 34 indicates that an additional. attribute, start time, must: be incorporated into the event, distribution task. Process 1.5 - 1.8 - Complete Tasks Additional attributes of the distribution task event are recorded in Figure 35. These include the data elements finish time, elapsed time, and quantity of cases all of which will eventually be used in the computation of productivity statistics. 187 fiat-def W marmwnoN “Imam umber {AbK EMFIJ'EE I -—0 Name. I! n: l “46ml 0": Work 0—” WORK Area Coda AREA n l , a of UkflmaELflMDhl ‘1'. whd-C»—- M m °—" 1X94 An” (ad'- WV! N6. 1YPE 135k 4O--‘ AV4C’ area-5%» ”at ompplue Area. Nan-3. WT“ '— L PAL-Lé'r Area. Caulc. ' 1cm PICJLING Area. cosr 1th <.--* Tm Code. VMT‘AL. GENTLE. rm made. 71041510 I 1 I AND 1 l OfltF’PlNé’ Vfiwisc>—i- 4 139k. I Cb: Chat Tm 6““ LOOOE M ed., Hang, case. W 1?:NL 1'an H r ‘5?! (bag. l2-k~$ifl7 we. TAOIL FIGURE 33 E-R Diagram of Information Required By Process 1.3 " Assign Employees to Tasks " 12* 1'7?!— I ‘flxfic 188 Type Home. T 1ack . ‘T 2 . I' n 91:121me ""0 7'» 17H9KL TYPE __0 Un'rf? a? 11m Aflovved I aahagany of h h FEfiTFflEHATuDhl 11¢>KI Y 1 5m 139k. 1n"“’ W hb. FIGURE 34 WORK AZEA E—R Diagram of Information Required By Process 1.4 " Mark Time That Task Begins " 189 1&92. ‘itfl; ’51"- “Type Code. C' 213-D. I T "0- “m - ' T09! Type. I T mwnon @21ch P+ion _ 1"!“ 154a: '--<>Lbfl+5 cfi"1hne' uuuzerunufia FIGURE 35 E-R Diagram of Information Required For Processes 1.5 — 1.8 " Complete Tasks " 190 Process 1.9 - Prepare Receiving Report The preparation of the receiving report is an integral part of the event, finished goods supplier shipment. This is indicated in Figure 36 by the addition of the receiving report no. attribute. Process 1.10 - Perform Consistency Check The process illustrated in Figure 37 necessitates comparing the quantity of finished goods inventory received with the quantity of the finished goods noted on the distribution task recorders. This is essentially a confirmation of similar attributes from independent events. Hence no relationship links the event distribution task to the other two entities. Process 1.11 - Record Employee Departure A. model of the data required to complete the timekeeping function is shown in Figure 38. An additional attribute, departure time, has been added to the event employee service. Process 1.12 - Compare Employee Times with Daily Task Recorders Once again, this process represents :3 confirmation stage in which the attributes of two different events are being compared. In this situation (Figure 39) the elapsed time shown on the task recorder is compared with the total time shown on the clock cards. 191 FINI‘bHEP W &. 6009:: mm Gale..- ¢MRu>EJL "’{>Ch*¢- r1 Fumes-0&0 rascal-b cacxzzs *Tbflc IL'NVENIDZ‘I I m {flupnuurf 4WDMH+hi OP :iufinemfl Line ”0' am 1 . . , Fumbuso 1::de M " @0095 0.3.; ' line HBO" auPPuerz, PP'“5 swupmnemrf \OMSIB val FIGURE 36 E-R Diagram of Information Required For Process 1.9 " Prepare Receiving Report " 192 Fume-Hep shamans“: 3H: PMEH‘I’ M, 1m... 1 line w. . 696%.: 5' ' 1 a M9924, Gandp WM Daria. HMMFH’. DnfitfleuxfkaQ FIGURE 37 E—R Diagram of Information Required By Process 1.10 "Perform Consistency Check” 193 1.0. ".’hkx VMAEI§¥XASi5 .Arrfiwal Cl¥flr+tu1b 12flbJ ‘:'Efli --4>fikun6L 'fiflne, ’flhnet ’fiwuy . W93 EhMHuyvsfé r1 ZiflaVKJE | (xvi Cendcr' Cadello cuntrr Cptfl' ’ W22 BOG “3 Name. FIGURE 38 E-R Diagram of Information Required By Process 1.11 " Record Employee Departure 194 Cad¢.Na __.‘IJ2 No. I-I—ONW fibrt FIGURE 39 E-R Diagram of Information Required By Process 1.12 " Compare Employee Times With Daily Task Recorders " 14" Horne. 13’“ tkh i Ebaandcr FE» Ewuac~E£L 1’““‘ FIGURE 40 E-R Diagram of Information Required By Process 2.1 " Verification and Edit " 195 Process 2.1 - Verification and Edit In this process a verification of employee identification and a task recorder sequence check is carried out in preparation for later performance and payroll runs. Information relating to these procedures is shown in Figure 40. Process 2.2 - Update Files An update of the employee file is carried out here and includes such items as employee additions and deletions, deduction changes, or rate changes. In addition, the employee time records arriving from the verification process provide details about employees' hours for the cmrrent period. These details will be manipulated and become part of the employee record. At the same time the payroll expenses are distributed to the work area and cost centers which incurred them. The view of the information needed to perform this process is shown in Figure 41. It can be noted that several more attributes than were previously recorded are now required. Process 2.3 - Prepare Payroll Reports and Print Checks Using information from the employee file and the authorized payments lists, this process models the information necessary to complete the payroll function. The event, cash disbursement is related to the resource cash and to the participants of the linkul 1me III a?~¢uNp data! any “Inez MflHZE+KXA§E E%MWHDfEEE 'éfflEkflCIE VWOEHL AQIZEUK MP F%¥Hau05poum was gonna mmsosmumz: H sumo nufi3 woumfioomm< mafismcoaumawm was mmfiufiuam msu mo Emuwmfin mum we mam—OHM Otis!“ 2‘s. . ‘3 G2 39‘s F 306. ~N .Stlgn \N “03. 08°.- 38 I 76¢§9286 dzafifiv o - 204 TABLE 6 Table Illustrating Attributes of Entity and Relationship Sets for Case 1 "The Warehouse Labor and Productivity System" Entity or Relationship Set Attributes Entities: FINISHED GOODS INVENTORY Finished Goods Inventory No. FINISHED GOODS PURCHASE ORDER Purchase Order No., Order Date. FINISHED GOODS SUPPLIER Shipment Identification No., SHIPMENT Shipping Date, Receiving Report No. SUPPLIER Supplier No., Supplier Name, Supplier Address. VENDOR Vendor No., Vendor Name, Vendor Address. PURCHASING AGENT Signing Authority I.D., Name. COST CENTER Cost Center Code No., Cost Center Name, Cost Center Address, Cost Center Type. PFW Cost Center Code No., Branch Name, Address, Manager, Budgeted Productivity Efficiency, Budgeted Productivity Effectiveness, Actual Labor Hours Accumulated, Actual Labor Cost Accumulated, Budgeted Labor Cost, Actual Productivity Efficiency, Actual Productivity Effectiveness. RDC Cost Center Code No., Branch Name, Address, Manager, Budgeted Productivity Efficiency, Budgeted Productivity Effectiveness, Actual Labor Hours Accumulated, Actual Labor Cost Accumulated, Budgeted Labor Cost, Actual Productivity Efficiency, Actual Productivity Effectiveness. TABLE 6 cont'd. 205 Entity or Relationship Set Attributes RECEIVING AGENT LDC WAREHOUSE EMPLOYEE SERVICE WAREHOUSE EMPLOYEE DISTRIBUTION TASK DISTRIBUTION TASK TYPE FULL PALLET TASK PARTIAL PALLET TASK LOOSE CASE TASK Signing Authority I.D., Name. Cost Center Code No., Branch Name, Address, Manager, Budgeted Productivity Efficiency, Budgeted Productivity Effectiveness, Actual Labor Hours Accumulated, Actual Labor Cost Accumulated, Budgeted Labor Cost, Actual Productivity Efficiency, Actual Producitivty Effectiveness. Arrival Time, Departure Time, Total Time. I.D. No., Name, Address, Phone No., Social Security No., Deductions Code, Exemption Code, Pay Rate, Paid Vacation Allowed, Sick leave Allowed, Hours Worked, Gross Earnings, Net Earnings, Tax Withholdings. Task Recorder No., Start Time, Finish Time, Quantity of Cases, Elapsed Time, Task Type Code. Task Type Code, Task Type Name, Task Type Description, Standardized Time Units, Quantity of Cases, Elapsed Time, Equivalent Cases, Units of Time Allowed. Task Type Code, Standardized Time Units. Task Type Code, Standardized Time Units. Task Type Code, Standardized Time Units. TABLE 6 cont'd. 206 Entity or Relationship Set Attributes DAMAGED CASE TASK WORK AREA RECEIVING AND STORING PICKING PACKING AND SHIPPING CASH CASH DISBURSEMENT DISBURSEMENT AGENT GOVERNMENT AGENT Relationships: FINISHED GOODS INVENTORY - FINSIHED GOODS SUPPLIER SHIPMENT Task Type Code, Standardized Time Units. Work Area Code, Work Area Name. Work Area Code, Budgeted Productivity Efficiency, Budgeted Productivity Effectiveness, Actual Productivity Efficiency, Actual Productivity Effectiveness, Budget Labor Cost, Actual Labor Hours Accumulated, Actual Labor Cost Accumulated. Work Area Code, Budgeted Productivity Efficiency, Budgeted Productivity Effectiveness, Actual Productivity Efficiency, Actual Productivity Effectiveness, Budget Labor Cost, Actual Labor Hours Accumulated, Actual Labor Cost Accumulated. Work Area Code, Budgeted Productivity Efficiency, Budgeted Productivity Effectiveness, Actual Productivity Efficiency, Actual Productivity Effectiveness, Budget Labor Cost, Actual Labor Hours Accumulated, Actual Labor Cost Accumulated. Bank Account No., Balance on Hand. Voucher No., Check No., Amount, Earnings Record No., Payroll Register No., Tax Register No. Signing Authority I.D. Agent Name, Agent Address, Amount Owing. Finished Goods Inventory No., Shipment Identification No., Shipment Line No., Quantity of Cases. TABLE 6 cont'd. 207 Entity or Relationship Set Attributes FINISHED GOODS INVENTORY - FINISHED GOODS PURCHASE ORDER FINISHED GOODS PURCHASE ORDER - FINISHED GOODS SUPPLIER SHIPMENT FINISHED GOODS PURCHASE ORDER - SUPPLIER - PURCHASING AGENT FINISHED GOODS SUPPLIER SHIPMENT - SUPPLIER - RECEIVING AGENT WAREHOUSE EMPLOYEE SERVICE - WAREHOUSE EMPLOYEE - COST CENTER DISTRIBUTION TASK - DISTRIBUTION TASK TYPE DISTRIBUTION TASK - WAREHOUSE EMPLOYEE - WORK AREA WORK AREA - COST CENTER CASH - CASH DISBURSEMENT CASH DISBURSEMENT - WAREHOUSE EMPLOYEE - DISBURSEMENT AGENT CASH DISBURSEMENT - GOVERNMENT AGENT - DISBURSEMENT AGENT RECEIVING AGENT - COST CENTER WAREHOUSE EMPLOYEE SERVICE - CASH DISBURSEMENTS Finished Goods Inventory No., Purchase Order No., Order Line No., Quantity of Cases. Purchase Order No., Shipment Identification No. Purchase Order No., Supplier No., Signing Authority I.D. Shipment Identification No., Supplier No., Signing Authority I.D. Arrival Time, I.D. No., Cost Center Code. Task Recorder No., Task Type Code. Task Recorder No., I.D. No., Work Area Code. Work Area Code, Cost Center Code. Bank Account No., Voucher No. Voucher No., I.D. No., Signing Authority I.D. Voucher No., Agent Name, Signing Authority I.D. Signing Authority I.D., Cost Centre Code. Arrival Time, Voucher No. 208 1.3 - 1.8 of Figure 29 and 2) Process 2.6 of Figure 30. While it would be a straightforward task to define each data element for every process of Figures 29 and 30, these two examples illustrate well the nature of the data element dictionary. Furthermore, the examples chosen contain the elements necessary to process the productivity and cost reports that will be described shortly. Talfile 7 begins by describing the daily task recorder data flow (that is, the output of Process 1.3 of Figure 29). Each of the three components (H? the daily task recorder are decomposed until finally a primitive (non-decomposable) set of data elements emerge. For example the component Task Recorder No. is seen to consist of a Cost Center Code and a Task Recorder Code. Within the former code is further information relating to Cost Center Type which itself is a collection of three additional components. When each of these components is further defined, a set of elements emerge which cannot be partitioned. These represent the primitive data elements. This procedure continues until each data flow is decomposed into its data elements, and each element is defined. TLable 8 lists the data dictionary definitions which make up the data flows and data files of Process 2.6 in Figure 30. The primitive elements which emerge from this procedure are those that will be used by the data base to produce the productivity and cost reports that were requested by the various distribution executives during the requirement analysis stage. 209 TABLE 7 Data Element Dictionary Entries for Data Flows Which Take Place when Employees Handle Finished Goods (An Ilustration of Processes 1.3-1.8 of Figure 29) Task Recorder N0. + Warehouse Employee I.D. No. + Warehouse Employee Name Daily Task Recorder Task Recorder No. Cost center code + Task Recorder Code Cost Center Type PFW RDC LDC * Production Facility Warehouse* * Regional Distribution Center* * Local Distribution Center* PFW = A Two digit code 01-09 indicating location = Ibl‘ 02 *Detroit* *Atlanta* * * L”O9J Los Angeles A two digit code 10-49 indicating location = Fllq RDC = *Lansing* 22 RH A two digit code 50-99 indicating location *Montgomery* *San Francisco* LDC = = ”SlI O *Houghton* 64 *Dothan* *Santa Cruz* .821 TABLE 7 cont'd. Task Recorder Code Warehouse Employee I.D. No. Warehouse Employee Name Open Daily Task Recorder Work Area Code Task Type Code Start Time Closed Daily Task Recorder Finish Time Elapsed Time Quantity of Cases Where 210 A four digit number 0001-9999 A six digit code 00001-99999 unique to each employee A three part identifier including first name, middle initial and last name. Daily Task Recorder + Work Area Code + Task Type Code + Start Time A single letter A, B or C indicating work function A *Receiving and Storing* B *Picking* C *Packing and Shipping* A single digit code 1-4 indicating task type 1 *Full Pallets Task* 2 *Partial Pallets Task* 3 *Loose Cases Task* 4 *Damaged Cases Task* Time/Day/Month/Year a Task is Begun Open Daily Task Recorder + Finish Time + Elapsed Time + Quantity of Cases Time/Day/Month/Year a Task is Completed Finish Time minus Start Time in Minutes Actual Count of cases handled by employee is equivalent to either or ** comments 211 TABLE 8 Data Element Dictionary Entries for Data Flows Which Take Place when Transaction Details are Transformed into Productivity and Cost Reports (An illustration of Process 2.6 of Figure 30) Transaction Detail File = {Task Recorder No.} + Work Area Code + Task Type Code + Elapsed Time + Quantity of Cases + Equivalent Case Factor + Standard Time Units + Equivalent cases + Units of Time Allowed Task Recorder No. I Work Area Code definitions of each of these data Task Code I components and data elements were Elapsed Time shown in Table 5 Quantity of Cases I 4 Equivalent Case Factor = The equivalent task time required to handle 1 case compared to a standard equivalent task time of 1 time unit per case in receiving and storing full pallets Standardized Time Units = Units of time allowed to handle a case Equivalent Cases ‘ = Quantity of Cases multiplied by Equivalent Case Factor Units of Time Allowed = Quantity of Cases Multiplied by Standardized Time Units Productivity Effectiveness = Cost Center Code+({Work Area Code}) (P.E.) Report Work Area Code P.E. Current week + P.E. Previous week + P.E. This-week-last-year + P.E. Current-month-Actual + P.E. Current-month-Budget + P.E. Last- Month-Actual + P.E. This-month—Last- Year-Actual Productivity Effectiveness= Quantity of Cases Divided by Total Units of Time Allowed TABLE 8 cont'd. Productivity Efficiency (P.Eff) Report Productivity Efficiency Work Area File Cost Center File Unit Labor Cost Report Unit Labor Cost Compensation per Man Hour Report Compensation per Man Hour Labor Cost Report Where 212 Cost Center Code+({Work Area Code}) Work Area Code P.Eff Current-week + P.Eff Previous- week + P.E. This-week-Last-Year + P.Eff Current-Month-Actual + P.E. Current-Month-Budget + P.Eff Last- Month-Actual + P.E. This-Month-Last- Year-Actual Total equivalent cases divided by elapsed time {Work Area Code + Work Area Name + {Actual Hours Accumulated + Actual Labor Cost Accumulated + Budgeted Productivity Effectiveness + Budgeted Productivity Efficiency + Actual Productivity Effectiveness + Actual Productivity Efficiency}} {Cost Center Code + Cost Center Name + {Actual Hours Accumulated + Actual Labor Cost Accumulated + Budgeted Productivity Effectiveness + Budgeted Productivity Efficiency + Actual Productivity Effectiveness + Actual Productivity Efficiency}} {Cost Center Code + Actual Labor Expense + Actual Quantity of Cases + Unit Labor Cost} Actual Labor Cost divided by Actual Quantity of Cases {Cost Center Code + Actual Labor Cost + Actual Hours + Compensation per Man Hour} Actual Labor Cost divided by Actual Hours Cost Center Code + Actual Labor Work Area Code Cost + Budgeted Labor Cost is equivalent to repetitions of either - or Optional key attribute 213 Although Tables 7 and 8 represent only a partial data element specification of the entire labor productivity and payroll system, they illustrate well the principles involved. The dictionary entries would be complete when each data flow and data file have been defined in terms of non-redundant primitive data elements. At a minimum, the following would make up a fully specified data dictionary [DeMarco, p. 36]. 1. For each data flow name, there is one dictionary entry. 2. For each data file, there is one dictionary entry. 3. For each component within a data flow or data file, there is one dictionary entry. 4. For each primitive, there is one dictionary entry unless the primitive is so obvious as to be self-defining. While the Tables present a logical partitioning of the relevant data flow or data file into its primitive data elements, each entry in actuality represents a separate definition 'which would be filed in the Data Dictionary in alphabetical order under the heading of Data Element name, Data Flow name, Data File name or Data Process name. Transform Descriptions A third output of the View Modeling Process of Figure 28 is a description of the policy governing the transformation of input into output data flows. For each process shown in the data flow diagrams, a data transformation or specification is made. DeMarco 214 (1979, p. 170-176) lists these four goals that such specifications should attempt to achieve. 1. Each specification should describe the rules governing the transformation of data flows arriving at the process into the data flows leaving it. 2. Each Specification should describe the policy underlying the transformation, not how to implement the policy. 3. Each specification should be non redundant. 4. Each Specification should maximize the possible degree of orthogonality (that is, the specification should use the minimum set of constructs to describe a process and these constructs should have a minimum of overlapping features). DeMarco (1979, Ch. 16) uses a language called Structured English to detail the workings of a data transformation. Table 9 and 10 provide two examples of how these procedures apply to the applications at hand. The illustrations focus on the examples highlighted in the last section. - namely' Processes 1.3-1.8 of Figure 29 and Process 2.6 of Figure 30. These applications were chosen because all the data elements necessary to describe the processes were defined earlier in Table 7 and 8. Hence, with the exception of certain transitive verbs (such as 'access', 'aggregate' and 'set' which indicate what is to be done), certain special words and phrases (such as 'for each', 'Otherwise', 'Case', 'If') and certain relational Operators (such as 'equal to' 'and' 'or'), all semantic information regarding objects and attributes 215 TABLE 9 Transform Descriptions of Process 1.3 to 1.8 Process 1.3 For each distribution task Assign warehouse employees to task types Issue daily task recorder Record the following Warehouse Employee I.D. No. Warehouse Employee Name Process 1.4 For each daily task recorder Record the following Work Area Code Task Type Code Start Time Process 1.5 - 1.8 For each open daily task recorder Record the following Quantity of cases Finish Time Elapsed Time 216 TABLE 10 Transformation of Transaction Detail Data into Productivity and Cost Reports - Process 2.6 Process 2.6 For each cost center For each work area For each task type Accumulate the quantity of cases Accumulate the elapsed time Set task productivity effectiveness equal to quantity of cases divided by elapsed time Set standard task productivity effectiveness equal to number of cases handled divided by units of time allowed Accumulate equivalent cases Set task productivity efficiency equal to equivalent cases divided by elapsed time Set standard task productivity efficiency equal to equivalent cases divided by total allowed time Accumulate quantity of cases handled in all tasks Accumulate elapsed time in all tasks Accumulate equivalent cases for all tasks Set work area productivity effectiveness equal to quantity of cases divided by elapsed time Set standard work area productivity effectiveness equal to quantity of cases divided by total allowed time Set work area productivity efficiency equal to equivalent cases divided by elapsed time Set standard work area productivity efficiency equal to equivalent cases divided by total allowed time Access the work area file Write work area productivity effectiveness and work area productivity efficiency ratios to work area file Prepare productivity reports Accumulate quantity of cases in all work areas Accumulate elapsed time in all work areas Accumulate elapsed time in all work areas Accumulate equivalent cases for all work areas Set cost center productivity effectiveness equal to quantity of cases divided by elapsed time Set standard cost center productivity effectiveness equal to quantity of cases divided by total allowed time Set cost center productivity efficiency equal to equivalent cases divided by elapsed time Set standard cost productivity efficiency equal to equivalent cases divided by total allowed time 217 TABLE 10 cont'd. Access the cost center file Write cost center productivity effectiveness and cost center productivity efficiency ratios to cost center file Retrieve cost center labor expense Set compensation per man hour equal to cost center labor expense divided by elapsed time Set unit labor costs equal to quantity of cases divided by cost center labor expense Write compensation per man hour and unit labor costs to cost center file Prepare productivity and cost reports 218 has been fully disclosed in the Data Dictionary. For example, definitions (fl? objects such as transactions detail file and cost center and definitions of attributes such as cost center code, elapsed time, and quantity of cases, can be found in the data dictionary. Thus the Structured English Specification can be coordinated with the Data Dictionary to ensure consistency. The New Logical Data Flow Diagram As illustrated in Figure 28, an additional output associated with the new modeling and modification procedure is the production of a new data flow diagram. This diagram incorporates the logical features of both existing needs and new requirements. With respect to the warehouse receiving and storing case, this involves combining the following figures. 1. Existing system - Figures 24 and 26 (Chapter 3) 2. New system - Figures 13 and 14 (or Figures 29 and 30) The new data flow diagrams are illustrated in Figures 47 and 48. Dashed lines denote the additions which have been added to the original system. A glance at the above mentioned figures reveals that the requirements of the new system contain all the information required to meet existing data needs. The difference in data requirements between the existing and new systems can be further illustrated by highlighting the additional data elements that must be captured 1»! the revised 219 FIGURE 47 Diagram 1: Receiving and Storing 220 FIGURE 48 Diagram 2: Payroll and Labor Distribution 221 information system. Such an illustration can be viewed in Figures 49 and 50. Here the solid lines represent existing processes whereas the dashed lines indicate the new processes. Summary of Case I At this point in the design process each of the elements of the View Modeling and Modification stage of Figure 28 are complete1 with respect to the Warehouse Labor Payroll and Productivity example. Attention now shifts to Case II - The Least Cost Distribution System. View Modeling and Modification - Case II In this section, the data requirements associated with the least cost distribution system are modeled. As in the previous case the procedure begins with the completed data flow diagrams that were produced for each application shown in Chapter 3. These diagrams consist of Figures lS-17 and for reader convenience are reproduced here as Figures 51, 52 and 53. Data Modeled Views Processes l, 2 and 3 - Receiving and Storing, Fill Order and Payroll and Labor Distribution 1 The exception. is the constraint specification. Admittedly this is an important practical consideration but plays only a minor role in this study. 222 1 290911ch avers/w DATA raw New 5v:>TEN\ meat-544T: nemmfa ‘1. was; mamas. / x \ \ .WW 02052. Amie-N \ ‘. IPMENT 3:9. 0’ r ‘ ' pA—rg - m5 _=- amen-r m—re / / .72" 1 '% 0F - — / ~ ‘3’ . g m! a: C» \ \ ~i- 79145.2 NUMBER. M42“ 1 Mar. 1% CODE / m m AQBA cope / AND graze 67m TIME \ "' W F‘N' 1" -- 83$?“ 0" \ A {>59 TIME. 2512:0021 I" r. I ”25- TIME FOhlétb’l'EN ‘I no AL. TIME. s m "I. ’VINQ '- ’.-1 K \ / - 52.2 . . \ / WEE; / ~ ‘ Ennwaznaifla . ‘\ COMPARE. \ lama/es. WtTH \wuw rm] W \ / FIGURE 49 \ -—— / Data Requirements for New and Existing System " Receiving and Storing " 223 amanNe ovsrw ’ NEW mare/w muzemeNre 9:1" HELP: IEEQUIZEMENT‘b ‘— VM32u=v' CALL-MW I"; 63 3:2I g II E ,3 Is I \_§/’ d““'llhfl '1axuu_ 1n~ul \\ ‘/ 1“- ' 15211425. TIME. \ __ / um FILEs - arm... 174A; 351‘ cur V W AREA caps. OFF DATE. : a '4 - - . ,.r .» 2; ""~ / ,. K Accoum :ur / \ ., ,, NUMBER. I m \ lQUMAEIHZ. ‘flZAhEbkfiflCTflBW cure \ DETAIL. / lAzewnmeaM-W LABOR - .. NT nameufl/ QNSHW \ ;// I ‘ DAZDIw ‘flME- unrr: ' _ ‘_ \\ ‘ " ' 11MB \ ANTI" as mass M ‘ on m cops wI . wganwr' coca. ' m 1 'wvase m / D n- - r m ——- / ~— casual FIEUZAfifli ‘VBAdZ EHJD’ WMZOLL. TA): azuamzars FIGURE 50 Data Requirements of New and Existing System " ayroll and Labor Distribution " 224 IZEIHEHJINHS AM¢D CflZflZHflGv Vendor fiififqih¥1+ ‘documner\ INVENTOIZ‘I FILE. W22, PW LE. eshéwaha 6% DATA ~ OBTAlN CUOTOMEQ CEEnAILfis ch23!"- bot-31in: alarm 'Vefi - than anal rcpar’f wng‘l'cé FIGURE 51 Diagram 0: Evaluate Distribution Alternatives 225 IZECIUFWE’ :1 : 6HIPMEN1‘5 HLE Ivuxhebund flflgfhumnrv ‘Flm . H mm ln~é3c¢b +7. Gonna?! F2636»? CCEHHD 1DH2MFF kc»: * ‘5"; 4' cos} by' ErnacflP hvrcafifvrv/ahmflflncflfiorr FIGURE 52 Diagram 4: Traffic Mm m d B nwue?mp ' carhyrficoat by Phaduef in M‘.’ Diagram 5: 226 FHLI. C057 GENT-.2. Fl Le. FIGURE 53 Inventory Carrying Costs byioca&hns 227 Figure 51 represents the parent diagranx and indicates six processes for which data models must be provided. Fortunately, most of the modeling associated with the first three processes of Figure 51 has already been accomplished in the previous case. These three processes are concerned with the determination of a variable warehousing (receiving and storing, picking, packing and shipping) labor cost by product by location. Such requirements are virtually the same as those uwdeled in the productivity case, with the following exception. In the productivity case, the desire was In) develOp productivity statistics for individual work areas and cost centers. No requirement existed for linking these statistics or costs to specific finished good inventory items. The present situation requires that a data link be constructed between the employee, the distribution task and the finished goods inventory. This is easily modeled by adding a relationship between the finished goods inventory and distribution task entities. Note that under this arrangement, an employee may record the handling of different finished goods inventory' numbers on the same distribution task recorder. Hence the attributes start time, finish time, elapsed time and quantity of cases are shown here on the relationship set instead of on the entity distributions task. This and other relevant features associated with Processes 1, 2 and 3 are modeled in Figure S4. 228 “1mm , AHV“ mm” ELUHH4DV££i ~——doCZq~u-huru.1‘n~t- .__—oflbha|‘“rn.' seamen. H t» g‘ | Co“ I can»; I \ H10. No . Ffifiehndlshad$ Invuwha1y bk» FtNitHED GuxDCHs Mow ,3 Wayman 1neuc FIGURE 54 E-R Diagram of Information Required By Processes 1,2 and 3 " Receiving and Storing, Fill Order, and Payroll.and Labor Distribution " 229 Process 4.1 - Match Invoice to Bill of Lading For each carrier invoice received, the HP company needs to verify that the goods were shipped and have been received. An E-R diagram of this process is shown in Figure 55. Process 4.2 - Compute Freight Costs Once the invoice and shipping documents have been matched, the process of allocating the costs of the shipment to specific inventory items can proceed. To accomplish this on an equitable basis, a combination of factors may be incorporated into the calculation. This will ensure a distribution of freight cost to the products in a manner similar to that used by the common carrieru. Modeling these requirements involves adding ‘new attributes to the finished goods inventory entity. For example, a National Motor Freight Classification (NMFC) code (a commodity code) has been incorporated to reflect the specific finished goods shipping characteristics. Appropriate processing of the above will result in a freight cost by product for the existing distribution patterns. However, the new proposal calls for the shipment of products from the PFW direct to the customer. This is a distribution strategy not previously followed by the HP company and one for which no actual freight data has been accumulated. Obtaining such information can be accomplished by incorporating a TPotential. Distribution an: My: mm». W3 230 13‘. Idunfi‘ihuflan No. gzazruwf I T FI'quap W3 Zak: 5|: acupcab Lgv’OCUfl‘, Fuiusnmu> ,,a4’fita.q3 P bureau “’0” fib'ppc‘ wmm , Paula 6|“ ”I’m Mm. '3 MP 1’ III—flake 'P n I fl ' I ‘GJHZEUEJI WH‘AJWHC. AGHEBPV €1-~‘*‘ 1,C2MH%¢- €536h° m I No. M”; J luame. FIGURE 55 Diagram of Information Required By Process 4.1 " Match Invoice to Bill of Lading " 231 Services" entity. As indicated by the multiple primary keys, this is a complex entity in the sense that the origins, destinations, NMFC codes and shipment sizes are multiplicative. The E-R diagram is illustrated in Figure 56. Process 5.1 - Obtain Unit Inventory Carrying Costs By Product By Location An important component of distribution costs is the inventory carrying cost associated with a product. Inventory carrying costs are a function of both product cost and stocking location characteristics. Hence, depending on where it is stocked, the same finished goods inventory item may have several different unit inventory carrying costs. Figure 57 provides an E-R model of this process. In the diagram it is the relationship between the finished good inventory entity and cost center entity that carries the needed attribute. Process 5.2 - Determine Average Inventory Carrying Cost By Product By Location In this process the average cost of inventory' on hand is required. When multiplied by the unit inventory carrying cost, an average inventory carrying cost on hand by product by location is obtained. The data model of this process is shown in Figure 58. 232 HFs'm‘sked Geode MW N" ~—-o muse. Code. “°(Anfi+ ‘Vkiuwna. FIGURE 56 E-R Diagram of Information Required By Process 4.2 "Compute Freight Costs" 233 my“? uuuwduwy 0 W50", a“ F." 0—4 Rum—tea coe‘xr ad vuahad , a: . Goa» 600” ‘ 65:41:42 Ca“ Ina-«bu, INVENTOZY No. nu» FIGURE 57 E-R Diagram of Information Required By Process 5.1 " Obtain Inventory Carrying Costs By Product By Location " Hhfiwud ’ Fennel-mp I 3;;7' (30+ I “" “w” 2:.” o... INVEBHRZ?Y . Lhwt Bk! Coo+ FIGURE 58 E-R Diagram of Information Required By Process 5.2 " Determine Average Inventory Carrying Cost By Product By Location " und+ bnundany . ° on was 095? Fbflflud Mr mucous: w_._.Coa|' W 509 M M 0051' N" P: at ‘ canvas: cad-.03.. . °—-+ Invem'ozv ' , Ubw+ and FIGURE 59 E-R Diagram of Information Required By Process 5.3 __‘:“ufiwnn¢ lid-nfiflhahbnih [=04wa 3H ‘_ Wu. EZZN25 ,__ofiwr&ual Ch+a §UUF%~£$*7 r—-O1fiiuwfih* ‘anv " Determine Average Inventory Carrying Cost By Product in Transit " 234 Process 5.3 - Determine Average Inventory Carrying Cost By Product in Transit In addition to the inventory carrying costs associated with a product at a particular storage location, carrying costs in transit must also be determined. The requirements of this process require a knowledge of the average inventory cost in transit at any one time and the unit inventory carrying cost. This is modeled in Figure 59. Process 6 - Obtain Customer Order Details As a final input into the least cost distribution choice decision, a knowledge of both past and future customer order details is required. Historical customer transactions provide details as to the past number of orders and sales by finished goods inventory item. In turn, management must provide estimates of how future order quantity and sales details will change with the proposed changes in delivery mode. The first part of this information set is determined by past events and is easily modeled. The latter part can be modeled by reviewing the Potential Sales entity and examining it for specific attributes related to forecasted customer sales. These details are modeled in Figure 60. Process 7 - Assemble Data Integration of all the data elements takes place at this stage and is discussed in the next heading. 23S “W‘: '5?“ FIGURE 60 E-R Diagram of Information Required for Process 6 "Obtain Customer Order Details" 236 Integration of Individual Views - Case 2 At this point in the design methodology, each of the processes of Figures 51, 52 and 53 have been modeled. The individual views developed in Figures 54-60 are now brought together to produce an integrated E-R diagram of the data requirements for Case II. This integration is shown in Figure 61. As in the previous case the attributes associated with each entity and relationship set can also be consolidated at this point. These are illustrated in Table 11. Data Element Dictionary, Transform Descriptions and New Logical Data Flow Diagram Completion of the View Modeling and Modification stage of Case II requires that the data element dictionary, transform descriptions and new logical data flow diagram be Specified. Since these procedures are relatively straightforward and have been previously illustrated in Case I, further work on these design features will be deferred to the View Analysis and Integration stage. View Modeling and Modification - Case III Figure 19 of Chapter 3 portrays the nature of the warehouse site selection decision. As discussed in Chapter 3, the data requirements associated with such strategic decisions differ from 237 FIGURE 61 E-R Diagram of the Entities and Relationships Associated with Case 2 "The Least Cost Distribution System" 238 TABLE 11 Table Illustrating Attributes of Entity and Relationship Sets for Case 2 "The Least Cost Distribution Decision" Entity or Relationship Set Attributes Entities: WAREHOUSE EMPLOYEE SERVICE Arrival Time, Departure Time, WAREHOUSE EMPLOYEE COST CENTER WORK AREA FINISHED GOODS INVENTORY DISTRIBUTION TASK DISTRIBUTION TASK TYPE FINISHED GOODS PURCHASE ORDER FINISHED GOODS SUPPLIER SHIPMENT DISTRIBUTION SERVICES CARRIER TRAFFIC AGENT POTENTIAL DISTRIBUTION SERVICES SALE FINISHED GOODS CUSTOMER ORDER Total Time I.D. No., Name, Pay Rate Cost Center Code No., Cost Center Name Work Area Code No., Work Area Name Finished Goods Inventory No., Unit Weight, Unit Volume, NMFC Code, Unit Cost, Selling Price, Forecast Selling Price, Quantity on Hand Task Recorder No. Task Type Code Purchase Order No., Order Date Shipment Identification No. Receiving Report No., Receiving Report Date, Total Weight, Origin Code, Destination Code, Shipment Date, Arrival Date Invoice No., Invoice Date, Cost Carrier Name, Carrier No. Signing Authority I.D., Name Origin Code, Destination Code, NMFC Code, Shipment Size, Cost Sales Invoice No., Date, Amount Sales Order No., Order Date TABLE 11 cont'd. 239 Entity or Relationship Set Attributes FINISHED GOODS CUSTOMER SHIPMENT CUSTOMER SALESMAN POTENTIAL SALES CASH RECEIPT Relationships: WAREHOUSE EMPLOYEE SERVICE - WAREHOUSE EMPLOYEE - COST CENTER COST CENTER - WORK AREA DISTRIBUTION TASK - WAREHOUSE EMPLOYEE - WORK AREA DISTRIBUTION TASK - FINISHED GOODS INVENTORY DISTRIBUTION TASK - DISTRIBUTION TASK TYPE FINISHED GOODS INVENTORY - FINISHED GOODS SUPPLIER SHIPMENT DISTRIBUTION SERVICES - FINISHED GOODS SUPPLIER SHIPMENT DISTRIBUTION SERVICES - CARRIER - TRAFFIC AGENT Shipment Identification No., Total Weight, Origin Code, Destination Code, Shipment Date, Arrival Date, Packing Slip No., Packing Slip Date Customer No., Customer Name, Customer Address, Ship-to- Location, Market Area Code, Total Sales to Date , No. of Orders, Balance Owing, Credit Limit Salesman No., Salesman Name, Accumulated Commissions Forecast No., Forecast Sales Value, Forecast No. of Orders Cash Receipt No., Amount, Date Arrival Time, I.D. No., Cost Center Code Cost Center Code, Work Area Code Task Recorder No., I.D. No., Work Area Code Task Recorder No., Finished Goods Inventory No., Start Time, Finish Time, Elapsed Time, Quantity of Cases Task Recorder No., Task Type Code Finished Goods Inventory No., Shipment Identification No., Shipment Line No., Quantity of Cases, Weight Invoice No., Shipment Identification No. Invoice No., Carrier Name, Signing Authority I.D. 240 TABLE 11 cont'd. Entity or Relationship Set Attributes FINISHED GOODS INVENTORY - COST CENTER POTENTIAL SALES - CUSTOMER POTENTIAL SALES - FINISHED GOODS INVENTORY FINISHED GOODS CUSTOMER ORDER - CUSTOMER - SALESMAN FINISHED GOODS CUSTOMER ORDER - FINISHED GOODS CUSTOMER SHIPMENT FINISHED GOODS INVENTORY - FINISHED GOODS CUSTOMER SHIPMENT FINISHED GOODS CUSTOMER SHIPMENT - DISTRIBUTION SERVICES CASH RECEIPTS - CUSTOMER - CASH RECEIPT AGENT CASH RECEIPT - SALE SALE - FINISHED GOODS INVENTORY SALE - FINISHED GOODS CUSTOMER SHIPMENTS FINISHED GOODS CUSTOMER ORDER - FINISHED GOODS INVENTORY Finished Goods Inventory No., Cost Center Code No., Unit Inventory Carrying Cost, Quantity on Hand Forecast No., Customer No. Forecast No., Finished Goods Inventory No., Forecast Quantity Sales Order No., Customer No., Salesman No. Customer Order No., Shipment Identification No. Finished Goods Inventory No., Shipment Identification No. Shipment Identification No., Invoice No. Cash Receipt No., Customer No., Signing Authority I.D. Cash Receipt No., Sale Invoice No. Invoice No., Finished Goods Inventory No., Quantity of Cases, Amount Invoice No., Shipment Identifica- tion No. Sales Order No., Finished Goods InventorypNo. 241 those of Case I and Case II in that information relating .to demographics, climatic conditions, community attitudes, local building codes and costs are not normally part of the corporate database. Data concerning these factors could be gathered and maintained internally but cost benefit considerations suggest that for non-routine decisions, it may be more cost effective to extract such information from published data sources. Despite these comments however, the more structured types of strategic decisions (of which the site selection is an example) will generally contain information requirements for which data may be regularly collected and maintained. In this case, such an example is transportation costs. As illustrated earlier in Table 4, inbound and outbound transportation costs are a critical factor in determining a warehousing site. In order to evaluate the potential of a new site, information must first be available on the current cost of servicing existing distribution linkages. Some of the modeling which took place in Case II is applicable here. However, that case dealt with a specific group of customers whose transportation needs were met by common carrier. In this case the focus changes to the entire freight system which includes common carriage, private carriage and U.P.S. Hence a model of the transportation costing system will be somewhat more complex in this situation. As in the previous cases the data models are constructed from the information provided in the data flow 242 diagrams. iFor reader convenience, the relevant data flow diagram (Figure 20 from Chapter 3) is reproduced as Figure 62. Data Modeled Views The E-R diagrams shown on the next few pages represent models of the data requirements for obtaining actual transportation costs from existing distribution linkages. Reflecting HP's freight policy, the data models include the three modes of freight; namely, private carriage, common carriage and U.P.S. For example, Figures 63 and 64 describe a model which first accumulates total trip costs, then allocates those costs to specific shipments on the basis of an inbound or outbound freight classification. Similar' models follow' for common carriage and U.P.S. shipments. Consolidation of the individual views takes place in Figure 69. In addition to the entities and relationships noted in the local views, a finished goods shipment type has been introduced to maintain data about the various private common and U.P.S. shipping modes. As in the previous cases the attributes associated with each entity and relationship set have been consolidated. These are illustrated in Table 12. This concludes the View Modeling and Modification stage for the new requirements outlined in Case I, II and III. Attention now turns to modeling the Existing System. 243 VEHICL-E. PM. E. ”mum an... m; Ina: finnflaca ._f JVuak MM divans a I FIGURE 62 (a) Context Diagram: Transportation Costs VEHICLE MILE new ' FIu. " lh‘ \M . .‘I “We” ckiuuns «unwanacur: «a: +9 -‘1fl-¢i flhwmnl a+hk,ca¢d WW “I? fuel FIGURE 62 (b) Figure 1: Obtain Inbound and Outbound Transportation Costs 244 : wumoo mane mwmfiuumu mum>wum oumvfiaomcoo : H.H mmoooum mm cmufisvmm cofiumahowcH mo Emuwman Mum no MMDOHm 24S :mfiufi casuas mucoaafinm ou umou many mumooHH<: .N.H oooooo m %m wmuaavom cowumauomcH mo amuwman Mum we MMDUHm .0“¢03§<.t$0$k?£0v LHHOAU .SPfiI 018 form!“ H ~33 0%.... $059.61 tag...“ 0:317 £20611 330 > Eng tilinol. «coon r EM“ .8. . .oz 33%? first: < taste“ o¥oAU cliftoce... 133R] 5.9.8.10 .D1Ru tiickisuoillj ones? Infill «Mafia as» Scum. \ 3:92.“. ‘ , . _ .oz . m 23L!“ c.9520 a....... at... 33c... 246 .. MESS mo :3 3 6335 sauna .. q.H mmoooum hm monasvom coaumsuowcH mo Emumman mum no MMDUHM 2a 5%.... In an 0°. H 952 Pipe ...Zusxtin a m3u<59tu30 no $000¢ c. F633» 3:92.“. ...Zmnu,‘ _ 9nu'Enq>kqpu: "" 1finnev Eiccrthyrs. ‘Anéubn p P ° Cos-f C2mn+rcfl Eécgfiwdgfi .ACKKJMbrfuflkl ilNhflfliTUmby cape: FILE. ”“9 EHBSEJ FlL—E .A~m~u*5c*bumfirj CI 5'“ 5 ‘ VENDOR purchase. Guerra ‘amflerfigflanwewr+ (:41i3r15 FIGURE 82 Diagram 1.2 : Order Receiving and Procurement 282 Hkieked ’fizuwfiir 0...... was: Han Gav- me cliff” C221Ln19~+ (:pndeal l°fiz°lvilo Pizraueua VWDGU‘ M l1=| | Acflvrn Order's v. L56! ‘3? .. ”VINE”, 25:02: my: ‘EIEEVWHEB 13""9 | FIGURE 83 Diagram 1.2.1 : Receiving and Storing Jinmnaa. ‘2 ' W Accounts EMSMW EEEHBVHENJE Cufmantw- W": Mm (kniafiéporf ed 255:. PAYABLE- Eénu++u:wzb Farfixnnancc. fiépor+a» Diagram 2: FIGURE 84 Accounting Information System I I . o ‘ l i I . I ’ ‘-“----- -”. ' 284 FIGURE 85 Diagram 2.2 : Payroll and Labor Distribution 285 flanflar labor coo:- conh-vl inH‘b‘ {ted ACQUDHON 02052 FILE. v"‘—;u: h I: 6N ” Y I5 EJVLPLOYE26 l-4' PREPARE cecawua ” _ MARK TIME 25m: ’l'HAT WK FIGURE 86 Diagram 1.2.1 : Receiving and Storing 286 .M W A ”mu numb“. lat; FHA! ' ED Iawvnamu . I I .24: -5 mm \\ | «cums ‘ 9“ M vars“. Q." on! vita, : \ ’~ A‘Hflu Aunueuuzza' I fault llnmnzargg__ : ' -NL- P.‘ 2., Gear | HI... mam ' m " ’ ~ Anna \ 1.5 mm caucus \ ‘ Wm \ 'wmmu 25mm un'd‘ ‘ ' .‘ cao+w, \\ hit“! ‘ . '"Phfl" «usher ${1' pr hay chunk. l ' ,'. ””7 W ' n1~£fif°7 mudfl+anao l 'frroduaf 1 :«Wm VNROLL- Ln?" ‘ Warm-r --- - - FIGURE 87 Diagram 2.2 : Payroll and Labor Distribution 287 and Procurement' as outputs and to 'Accounting Information System' and 'Payroll and Labor Distribution' as inputs. With the changes, the net inputs and outputs are balanced and a new set of logical data flow diagrams emerge. While not shown here, the process just described would be repeated for both Case II and Case III requirements. View Analysis and Integration - Procedural Phase Having declared the conceptual schema and new logical data flow data diagrams, attention turns to the second and final phase of View Analysis and Integration. Here, tranform descriptions describing the processes required to derive local views are carried out. In addition, changes which must be made to the data element dictionary as a result of the integration process are completed at this point. Each of these is now briefly described. Transform Descriptions The ultimate goal of the modeling process is to successfully present the user with his required information. At this point in the methodology, each user request has been (1) incorporated in a data flow diagram, (2) modeled with a local entity-relationship diagram and (3) integrated in the final schema. Now, from this final schema, the designer must be able to derive the local views associated with the user's request. Accomplishing this objective requires a set of procedural specifications called transform descriptions which. describe the 288 steps required to derive the desired information. Such specifications may be incorporated into a number of possible data sublanguages [Date, 1977]. One such language is called relational algebra, :3 relationally complete, simple, yet powerful language. Relational algebra applies operators such as SELECT, PROJECT, JOIN and others, to specific entity and relationship sets to produce a new set of relations containing the information required by the user. This information can be extracted from the derived relations and presented to the user in report form. Table 13 illustrates the use of the relational algebra3 language for one of the requirements outlined in Case I - The Warehouse Productivity System. Described here is a set of procedures which result in the production of productivity efficiency and effectiveness statistics for each task type in each work area and a summarized productivity statistic for each work area. Similar descriptions can be made for other processes. It is at this stage ‘where information. needed to produce distribution requirements (productivity reports, least cost alternative and warehouse site selection information) and accounting requirements (balance sheet, income statement: and payroll reports) are specified. An accounting transformation, for example, might include the selection, projection and summing of cash, fixed assets 3 Several commercial relational languages are available to perform these Operations. Examples are SEQUEL, QBE and a language for CP/M based microcomputers called dBASE II. 289 TABLE 13 Relational Algebra Example Illustrating Procedures For Obtaining Task and Work Area Productivity Reports Relational Algebra Instruction Comment USE WORK AREA - DISTRIBUTION TASK -EMPLOYEE SELECT WORK AREA - DISTRIBUTION TASK - EMPLOYEE WHERE WORK AREA = "NO." GIVING TEMPl JOIN TEMPl AND DISTRIBUTION TASK OVER TASK RECORDER NO. GIVING TEMP2 SELECT TEMP2 WHERE DISTRIBUTION TASK TYPE CODE = "NO." GIVING TEMP3 JOIN TEMP3 AND DISTRIBUTION TASK TYPE OVER DISTRIBUTION TASK TYPE CODE GIVING TEMP4 PROJECT TEMP4 OVER TASK RECORDER, QUANTITY OF CASES, ELAPSED TIME, UNITS OF TIME ALLOWED, EQUIVALENT CASES GIVING TEMPS SUM QUANTITY OF CASES, ELAPSED TIME, UNITS OF TIME ALLOWED, EQUIVALENT CASES FOR TASK RECORDER NO > 'xx' and < 'xxx' DO TASK PRODUCTIVITY CALCULATIONS AND STORE IN TSUMl (THIS COMMAND IMPLIES A SUBROUTINE OF PROGRAM STATEMENTS TO CALCULATE PRODUCTIVITY RATIOS) *Using the Relationship Set Employee, As 3 Starting Point *Selecting a Specific Work Area Code Number from this Relationship to Create a New Table called TEMPl *Joining the New Table TEMPl to the Entity Distribution Task to Obtain Information about the Distribution Tasks carried out in the Specific Work Area *Selecting out Specific Distribution Tasks (for example, the Full Pallet Task) *Combining the Attributes Associated with the Specific Distribution Tasks and Additional Information about those Tasks Contained in the Distribution Task Type Entity *Extracting those Attributes which are Required for Productivity Information and Placing these in a New Table - TEMPS *For each Work Area - Task Type Code Combination, Sum the Indicated Attributes for the Desired Time Period. *Perform the Calculations required to Obtain Actual and Standard Productivity Effectiveness and Efficiency for a Specific Distribution Task Type in a Specific Work Area. 290 TABLE 13 cont'd. Relational Algebra Instruction Comment (REPEAT FOR EACH TASK TYPE CODE) SUM TSUMI, TSUM2, TSUM3, TSUM4 OVER STANDARD PRODUCTIVITY EFFECTIVENESS, ACTUAL PRODUCTIVITY EFFECTIVENESS, STANDARD PRODUCTIVITY EFFICIENCY, ACTUAL PRODUCTIVITY EFFICIENCY AND STORE IN WSUM1 (REPEAT FOR EACH WORK AREA CODE) Note: OBTAINING PRODUCTIVITIES FOR COST CENTERS AND THE OVERALL COMPANY IS SIMILAR TO THE ABOVE. OBTAINING THESE WOULD SIMPLY INVOLVE STARTING WITH THE COST CENTER - WORK AREA RELATION- SHIP AND ENDING WITH AN ADDITIONAL 2 SUM STATEMENTS GIVING PRODUCTIVITIES FIRST BY COST CENTER (BY SUMMING THE WORK AREAS WITHIN A COST CENTER) THEN BY THE OVERALL COMPANY (BY SUMMING THE COST CENTERS WITHIN THE COMPANY). PORTION OF CONCEPTUAL SCHEMA USED IN *Sums the Productivities of the 4 Task Types (Full Pallet Task, Partial Pallet Task, Loose Case Task and Damaged Case Task) to obtain the Overall Productivities for the Specific Work Area THE TRANSFORM DESCRIPTION Distribution Task ‘Work ‘ Area Category of Distribution Task Type Employee 291 and customer balances to obtain total assets. The transform description component of the procedural phase of design is complete once all the local views required from both the new and existing system have been specified. Conclusion This chapter presented the results of modeling and integrating the user requirements developed in Chapter III. Modeling began by deriving an Entity-Relationship diagram for each process described in the logical data flow diagrams. Attributes were added to the entities and relationships according to the local data requirements. At the end of the three research cases a partial integration was made resulting in a diagram containing the configuration of entities and relationships required to support specific user requirements. In order to more fuly specify the processes under study, data element dictionary entries were recorded, transform descriptions were written and a new logical data flow diagram was drawn. Once the new and existing requirements were modeled, the analysis and integration phase began. The analysis of the local E-R diagrams removed redundancies and minimized problem areas encountered when different local views of similar entities prevailed. Integration of the new and existing requirements ensued. The conceptual schema resulting from. this integration contains the entities, relationships and attributes necessary to 292 meet the information needs of both accountants and several levels of distribution executives. The design phase concluded with examples of new logical data flow diagrams, transform descriptions and data element dictionary modifications which resulted from the final integration. Several accomplishments may 1x2 noted here. First, a methodology has been presented which systematically incorporates diverse information requirements into a single enterprise view. This means that data becomes a shared corporate resource removing the need for the redundant and inefficient ad hoc data gathering systems presently found in information systems such as distribution. Second, the disaggregated principle upon which the methodology is based means that the detail required by individuals in different functions and at different managerial levels, can be supported. Third, the umthodology is implementation independent meaning that it can be utilized with almost all of the existing database management systems in Operation. In the concluding chapter, these and other features relating to the methodology and its implications will be described more fully. 293 References Atzeni, P., C. Batini, V. DeAntonelli, M. Denzerini, F. Villanedi and B. Zanta (1982), "A Computer Aided Approach for Conceptual Data Base Design," in Automated Tools for Information System Design, (North Holland Publishing Co., 1982), pp. 85-106. Chen, P. (1976), "The Entity-Relationship Model - Toward a Unified View of Data," ACM Transactions on Data Base Systems, (March, 1976, pp. 9-36. Date, C.J. (1981), An Introduction to Database Systems, 3rd edition, (Addison-Wesley, 1981). DeMarco, T. (1979), Structured Analysis and System Specification, (Prentice-Hall, 1979). McCarthy, W.E. (1979), "An Entity-Relationship View of Accounting Models," The Accounting Review, (October, 1979), pp. 667-86. (1982), "The REA Accounting Model: A Generalized Framework for Accounting Systems in a Shared Data Environment," The Accounting Review, (July, 1982), pp. 554- 578. Teorey, T.J. and ThP. Fry (1982), Design of Database Structures, (Prentice-Hall, 1982). CHAPTER V SUMMARY The purpose of Chapter V is to present a summary of the dissertation, implications of the research and areas of possible future emphasis. The chapter begins with a brief summary of the motivation behind the study and the objectives associated with the research. The next sections discuss the methodologies used and the implications of the model for accountants and non-accountants. This is followed by a brief discussion that contains topics which the research did not address and some suggestions for future research. Summary of Research Motivation and Research Objective The motivation for this dissertation was sparked by four main factors. First, previous research exposed an imbalance between the demand for and supply of distribution information. Second, at least part of this information imbalance was a result of a dependence on traditional accounting constructs and ad hoc data gathering schemes. Third, recent advances in accounting data modeling and database design ‘methodologies suggested a set of procedures for overcoming such limitations. Fourth, previous work on accounting data models indicated that research integrating 294 295 accounting information systems with other large systems such as distribution would represent a significant step forward. With these issues in mind a study was defined whose purpose centered on improving the status of distribution accounting. Specifically, the objective was to describe and test a generalized methodological framework for analyzing, capturing and retrieving distribution information. Methodologies and Research Design.Used in the Study Several theories and methodologies were combined in the study to achieve the above objectives. First, the events approach to accounting provided the theoretical underpinning for capturing large amounts of disaggregated distribution data. Second, the Lum E}. 22*: approach to database design resulted in a methodological framework for dividing the project's design work into three distinct phases: Requirements Analysis, View Mbdeling and Modification, and View Analysis and Integration. Third, the three phases were Operationalized by the Structured Analysis and Entity- Relationship methodologies. The former provided a systematic way of proceeding through the Requirements Analysis phase to discover the nature and quantity of data needed to satisfy distribution decision-maker requirements. The latter modeled these requirements with the use of corresponding entity and relationship sets. Working from the outputs of Structured Analysis, View Modeling 296 derived a series of local E-R views which were subsequently integrated in the View Analysis phase to produce the final distribution accounting conceptual schema. To test the reasonableness of this combined. methodology' a research situation including three test cases and an existing system was described. The test cases were based on literature findings which indicated that information necessary for sound distribution operations and decision. making ability' was either unavailable from existing accounting systems or was being supplied by independent data gathering efforts. The existing system described a set of operations realistic enough to represent the general nature of physical distribution systems. The methodology was then applied to this research situation. In each case, the Structured Analysis approach was used to determine new and existing requirements. These findings ‘were subsequently passed on to the View’ Modeling phase where each individual or local view of the data was modeled by an E-R diagram. Integration of these views proceeded next resulting in a global view of the data model required to support the information needs of many different users. Finally, procedural specifications ‘were carried out to ensure the ability of the modeled system to produce the results outlined in the research situation. 297 Review of The Results and Business Implications A major thrust of the research was to develop a methodology which combined both the requirements analysis and the information analysis and definition aspects of design. While DeMarco's [1979] structured analysis approach satisfied the fomer, it was not an appealing set or procedures for the latter. Similarly, while the entity relationship approach represented a theoretically sound and pragmatic way to model data, it did not provide as systematic a set or procedures for obtaining data requirements as did the structured analysis approach. Consequently, the best features of the two approaches were combined into a single methodology. Emerging from this development is a design approach which enables the designer and user to systematically specify, model and integrate information requirements. Such a development should provide a contribution to conceptual database design. The problems involved in accounting for distribution activities provided a way to test the methodology in a pragmatic setting. Some observations concerning the results of this enquiry are now presented. The methodology was applied at three different levels: operational control, management control and strategic planning. In all three instances, it was found that the data requirements utilized by specific processes could be well described and modeled with the use of Data Flow and Entity-Relationship Diagrams. At the lower levels, it turned out that virtually all the data necessary 298 to meet the stated information requirements was or would be made available through the internal system. At the higher levels only a small portion of the data needs could be satisfied internally. At all levels, however, the data flow' diagrams were particularly useful in delineating the specific requirements associated with each process. If these requirements were deemed important enough to be captured internally, they became the basis for Entity- Relationship and attribute specification. Thus, a major conclusion of the study is that the methodology successfully describes and models the existing and required information needs of a large system. It was argued that the use of the E-R modeling approach would result in an integrated database system which could be shared by many and varied users. The conceptual schema which resulted from this study represents a consolidation of views which meets both the financial needs of accountants and the managerial needs of several levels of distribution executive. Thus another major conclusion of the study is that the methodology provides a shared data resource which satisfies the requirements of both accountants and non- accountants. The dynamic nature of business organizations requires a systems design that can be rapidly and efficiently updated. It was apparent that the partitioning characteristics of the methodology 299 provide a flexible and systematic approach to the problem of handling new or changing requirements. Not only does the nature of the logical data flow diagrams (DFD) make it easy to trace the effect of these updates throughout the system, but changes in the DFD's also can be used to immediately effect the local and global EéR data models. Thus a third conclusion is that the methodology provides for design flexibility and integrity. Without an actual field implementation of the methodology, generalizations about the result to potential users must be made cautiously. Nevertheless, the following comments are in order. AccountiggflImplications The use of a methodology which is independent of physical considerations means that the accountant should be able to take a more active role in the design of the corporate information system. This is true for two reasons. First, the methodology should be useful in providing accountants with a systematic and flexible way to obtain, update or delete user requirements. Use of the logical DFD approach ensures that the accountant's emphasis is properly placed on defining needs rather than on processing needs. Second, and perhaps more importantly, the methodology allows accountants to proceed beyond their generally accepted role as information provider to that of information modeler. Until the development of second generation data models, accountants were largely precluded 300 from the View Analysis and View Integration stage of design. Earlier models stress data structures and require familiarity with specific database management systems, such as hierarchical, network or relational. The complexity of these issues is such that few accountants can successfully proceed beyond the requirements specification phase. Consequently, accountants have found themselves increasingly giving way to other disciplines in the design and maintenance of the corporate information system. Use of the E-R model should help in overcoming this. Such second generation approaches emphasize data models rather than data structures and can be employed independent of the specific database management system in use. For accountants, particularly management accountants, who wish to regain a leadership role in the specification of their organization's information system, this development should be a welcome one. Use of the E-R data model permits the accounting information system to move from an independent system to an integral part of the database. However, it must be recognized that the use of a model which emphasizes entity and relationship sets as opposed to debit/credit conventions is a major departure from traditional practice. This is likely to be a more serious problem for accountants themselves who are accustomed to conventional accounting classifications than for non-accounting users who are more concerned with information content than double-entry bookkeeping. Nevertheless, the implication is that accountants 301 must be prepared to produce local views from the conceptual schema which correspond to the views of conventional accounting users. McCarthy [1982] illustrates that the process of deriving a traditional chart of account classifications from the schema is indeed desirable and feasible. Nevertheless, it should be stressed that use of the E-R model will likely require significant efforts on the part of the accountant to move between a conceptual world based on traditional accounting constructs to one based on E-R notions. Distribution Implications From the point of view of distribution personnel, the methodology described here overcomes many of the information problems described in earlier chapters. First, it represents a significant step forward in addressing Ray's criticism: What is still required is an underlying methodological base for producing distribution cost analysis of a meaningful nature [Ray, 1975, p. 85]. Combining the structured analysis approach for requirements analysis with the entity-relationship approach for requirements modeling results in a methodological framework for addressing a large variety of distribution cost problems. Second, the events theory upon which the methodology is based, means that large amounts of disaggregated data may be stored and utilized in the decision and control models utilized by the distribution manager. Decisions must, of course, be exercised on 302 the amount of data captured and the length of time such data is maintained in disaggregated form. However, these decisions are no longer based solely on the needs of the traditional accounting system; decisions to aggregate data are made 'knowingly' and in conjunction with distribution personnel. Third, the necessity of maintaining independent and ad hoc data gathering systems to capture distribution information can be largely eliminated. The disaggregated data collection procedures encouraged and the integration of different users' needs provided by the methodology results in an information system which supports both distribution and non-distribution requirements. As was noted earlier, however, the ability of the system to satisfy user requirements depends in part on the cost/benefit relationship attached to the collection of individual data elements. Consequently, it is apparent that the system described, because of its predominant reliance on past events, is more effective in meeting lower and middle level management requirements than the broad spectrum of requirements associated with top level management. Finally, since the methodology is pictoral rather than narrative in nature, distribution personnel can become more active participants in the design of their information requirements. Both the data flow diagrams and the entity-relationship diagrams provide a systematic and clear portrayal of data inputs, outputs and files. As the design of the system progresses, the distribution executive 303 can easily confirm or modify the data flows and data models simply by examining the latest diagram. As a result of the interaction taking place between designer, accountant and distribution executive, a set of diagrams emerge, supported by' appropriate definitions and processes, which should provide a realistic representation of distribution information requirements. Issues Not Addressed By The Dissertation This section deals briefly with issues that bear on the methodology but were not explicitly discussed in the dissertation. A complete process of database design includes not only the phases of Requirements Analysis, View Analysis and View Integration but also the phases of hnplementation and Physical Database Design. [Lum ‘35 .El" 1979.] While these latter two are critical components, they deal more specifically with the actual management of the database and with particular software and hardware features. In view of the study's emphasis on the logical aspects of database design rather than the physical, these issues were not elaborated. A second issue not specifically addressed is the cost-benefit aspects associated with information system design. Throughout the study an implicit assumption existed that the needs described by various accounting and distribution personnel were of sufficient value to be incorporated into the design. No attempt at determining the economics of the information was carried out. 304 Finally, issues regarding allocations of costs to specific responsibility centers, segments or products were only partially addressed. While Appendix A in Chapter III represents an example of how costs might be distributed for evaluation. and control purposes, cost allocation issues throughout the rest of the study were largely ignored. Conclusions and Directions for Future'Research This dissertation was undertaken to determine whether a methodology based on an events accounting approach and incorporating recent develOpments from database design could be used to model a large distribution accounting system. To the extent that it is possible to extrapolate the findings from a case study to field results, it seems justifiable to conclude that an entity-relationship approach to distribution accounting can be used to satisfy the information requirements of both accountants and distribution decision makers. Much work yet needs to be done, however, and several research directions are suggested by the dissertation. First, implementation of such a distribution accounting data model in an actual organizational setting would quickly expose its strengths and weaknesses. Hence a logical extension of this work would be to field test the methodology in a business setting. Second, research (”I the cost-benefit or economics of collecting and maintaining data is an issue of fundamental 305 importance to accountants and database designers. While the theoretical literature offers guidance on determining the value of additional information more work must be carried out before these issues are settled in a quick and practical manner. Third, research on the appropriate or optimum number of entity generalizations and associations in the conceptual schema could be pursued. Presently, no specific guidelines, other than the modelers' intuition, exist as to the level of redundancy which should be permitted in the model. The last two directions coincide with the suggestions made by McCarthy [1982]. Numerous behavioural issues present themselves in designing a system based on. the E-R approach” It should be possible to design a research study along the lines of Benbasat and Dexter [1979] to explore the behavioral differences between traditional accounting approaches and those encountered in the E‘R approach. Finally, while the present study was concerned with the functional area of distribution in a manufacturing environment, further research using the methodology could be carried out into such diverse areas as non-business accounting, social accounting, marketing and production accounting. 306 References Benbassat, I. and A.S. Dexter (1979), "Value and Events Approaches to Accounting: An Experimental Evaluation", The Accounting Review, (October, 1979), pp. 735-49). DeMarco, T. (1979), Structured Analysis and System Specification, (Prentice-Hall, Inc., 1979). Lum, V., S. Ghosh, M. Schkolnik, D. Jefferson, 8. Su, J. Fry, T. Teorey and B. Yao (1979), "1978 New Orleans Data Base Design Workshop Report," Research Report RJ2554 (IBM Research Laboratories, San Jose, CA, July, 1979. McCarthy, W.E. (1982), "The REA Accounting Model: A.Ckneralized Framework for Accounting Systems in a Shared Data Environment", The Accounting Review (July, 1982), pp. 554- 578. Ray, D. (1975), "Distribution Costing - The Current State of the Art," International Journal of Physical Distribution, Vol. 6, NO. 2, [1975], pp. 73-1070 Teorey, T. and J. Fry (1982), Design of Database Structures, (Prentice-Hall, 1982). B IBL IOGRAPHY 307 BIBLIOGRAPHY American Accounting Association (1969), "Report of Committee on Managerial Decision Models," The Accounting Review, (Supplementary 1969), pp. 43-76. Anthony, R. (1970), Management Accounting, Principles, revised edition, (Richard D. Irwin, 1970). Atzeni, P., C. Batini, V. DeAntonelli, M. Denzerini, F. Villanedi and B. Zanta (1982), "A Computer Aided Approach for Conceptual Data Base Design," in Automated Tools for Information System Design, (North Holland Publishing Co., 1982), pp. 85-106. Benbassat, I. and A.S. Dexter (1979), "Value and Events Approaches to Accounting: An Experimental Evaluation," The Accounting Review, (October, 1979), pp. 735-49. Bowersox, D.J. (1973), Dynamic Simulation of Physical Distribution Systems, (East Lansing, Michigan: Division of Research, Michigan State University, 1973). (1978), Logistical Management, 2nd edition, (MacMillan Publishing Co. Inc., New York, N.Y. 1978). Bream, R.E. and R. Galer (1974), A National Survey of Physical Distribution Management, (Whitehead and Partners, 1974). Bubenko, J.A. (1977), "IAM: .An Inferential Abstract Modeling Approach to the Design of Conceptual Schema, "ACM-SIGMOD International Conference on Management of Data, (August, 1977), pp. 62-75. (1977), "The Temporal Dimensions in Information Modeling," in G.M. Nijssen, ed., Architecture and Models in Data Base Management Systems, (North Holland Publishing Company, 1977), pp. 93-118. Buffs, E. (1976), Operations Management: The Management of Productive Systems, (Wiley, 1976). Carter, D.M., H.L. Gibson and R.A. Rademacher (1975), A Study of Critical Factors in Management Information Systems for the U.S. Air Force, (Colorado State University, 1975). 308 Chen, P. (1976), "The Entity-Relationship Model - Toward a Unified View of Data," ACM Transaction on Data Base Systems, (March, 1976), pp. 9-36. Christopher, M. and D. Ray (1976), Costing in Distribution: Problems and Procedures, (MCB Books, 1976). CODASYL Programming Language Committee (1971), Data Base Task Group Report (Association for Computing Machinery, 1971). Codd, E.F. (1970), "A Relational Model of Data for Large Shared Data Banks," Communications for the ACM (June 1970), pp. 377- 387. (1972a), "Further Normalization of the Data Base Relational Model," in R. Rustin, ed., Data Base Systems (1972b), "Relational Completeness of Data Base Sublanguages," in R. Rustin, ed., Data Base Systems (Prentice- Hall, 1972), pp. 65-98. Colantoni, C.S., R.P. Manes and A.B. Whinston (1971), "A Unified Approach to the Theory of Accounting and Information Systems," The Accounting Review, (January 1971), pp. 90-102. Constantin, J.A., RAD. Anderson and R.E. Jerman (1977), ”View of Physical. Distribution 'Managers," Business Horizons, (April, Cox, R., quoted in R. Moyer (1972), A Social Perspective, (John Wiley and Sons, 1972), p. 52. Date, C.J. (1981), An Introduction to Database Systems, 3rd edition, (Addison-Wesley, 1981). Davis, J. (1977), ”Distribution Systems Analysis," International Journal of Physical Distribution Management and Materials Management, Vol. 8, No. 2, (1977), pp. 74-88. DeMarco, T. (1979), Structured Analysis and System Specification, (Prentice-Hall Inc., 1979). Drucker, P. (1962), "The Economy's Dark Continent," Fortune, (April, 1972). Eaves, B.C. (1966), "Operational Axiomatic Accounting Mechanics," The Accounting Review, (July, 1966), pp. 426-442. 309 Everest, G.C. and R. Weber (1977), "A Relational Approach to Accounting MOdels,” The Accounting Review, (April, 1977), pp. 340-359. (1974), "The Objectives of Data Base Management," in Julius T. Tou, ed., Information Systems: COINS IV (Plenum, 1974), pp. 1-350 Flaks, M. (1963), "Total Cost Approach to Physical Distribution," Business Management, Vol. 24 (August, 1963), pp. 55-61. Fry, J.P., T.J. Teorey, D.A. DeSmith and L.B. Oberlander (1978), Survey of State: of the Art Database Administration Tools: Survey Results and Evaluation, Technical Report, DSRG 78 DE 14.2 Division of Research, Graduate School of Business Administration, University of Michigan, Ann Arbor, MI, 1978. Gane, C. and T. Sarson (1979), Structured Systems Analysis: Tools and Techniques, (Prentice-Hall, 1979). Gessford, .J. (1980), Modern Information Systems, (Addison-Wesley, 1980). Gold, B. (1980), "Practical Productivity Analysis for Management Accountants," Management Accounting, (May, 1980), pp. 31-44. (1979) , Productivity, Technology and Capital: Economic Analysis, Managerial Strategies and Government Policies, (Heath-Lexington, 1979). Graham, A.S. (1978), quoted in ”Computers: Making Waves in Distribution" in Contemporary Physical Distribution by J .C . Johnson, (Petroleum Publishing Company, 1978), pp. 62-66. Haseman, W.D. (1977), Introduction to Data Management, Richard D. Irwin, 1977). and Whinston, A.B.(l976), "Design of A Multidimensional Accounting System," The Accounting Review, (January, 1976), pp. 65-79. Horngren, C.T. (1982), Cost Accounting: A Managerial Emphasis, 5th ed., (Prentice-Hall, 1977). House, R.J. and B.R. Jackson (1978), "Trends in Computer Applications: A Survey," in Contempprary Physical Distribution by JRC. Johnson, (Petroleum Publishing Company, 1978), pp. 62-66. and J.J. Karranbauer (1978), ”Logistics Systems Modelling," International Journal of Physical Distribution and Materials Management, Vol. 8, No. 4, (1978), pp. 189-199. 310 Ijiri, Y. (1967), The Foundation of Accounting Measurement, (Prentice-Hall, 1967). Johnson, 0. (1970), "Toward an 'Events' Theory of Accounting," The Accounting Review, (October, 1970), pp. 641-53. Kearney, A.T. (1978), Measuring Productivity in Physical Distribution, (Chigago: National Council. of Physical Distribution Management, Chicago, IL, 1978). Keen, P. and M. Scott Morton (1978), Decision Support Systems: An Organizational Perspective, (Addison-Wesley, 1978). Lai, A. and B. LaLonde (1974), "An Analysis and Evaluation of Data Base Requirements for Physical Distribution Studies,” Physical Distribution Management, (Volume 4, No. 4, 1974), pp. 217- 248. LaLonde, B.J. and P.H. Zinszer (1976), Customer Service: Meaning and Measurement, (Chicago: National Council of Physical Distribution Management, Chicago, IL, 1976). J.R. Grabner and J.F. Robeson (1970), "Integrated Distribution Systems: A. Management Perspective," International Journal of Physical Distribution, (October, 1970), pp. 133-139. Lambert, D.M. (1978), The Distribution Channels Decision, (New York: The National Association of Accountants, and Hamilton, Ontario: The Society of Management Accountants of Canada, 1978). (1976), The Development of an Inventory Carrying Costing Methodology: A Study of the Costs Associated with Holding Inventory, (Chicago, National Council of Physical Distribution Management, 1976). Lambert, D.M. and J.R. Stock (1982), Strategic Phy§ical Distribution Management, (Richard D. Irwin, Homewood, IL), 1982. J.F. Robeson and J.R. Stock (1978), "An Appraisal of the Integrated Physical Distribution Management Concept," International Journal of Physical Distribution and Materials Management, Volume 9, Number 1, (1978), p. 74. LeKashman, R. and J.F. Stolle (1965), "The Total Cost Approach Distribution," Business Horizons, Vol. 8 (Winter, 1965), pp. 33-46. 311 Lieberman, A.Z. and A.B. Whinston (1975), "A Structuring of an Events-Accounting Information System," The Accounting Review, (April, 1975), pp. 246-258. Lum, V., S. Ghosh, M. Schkolnik, D. Jefferson, S. Su, J. Fry, T. Teorey and B. Yao (1979), "1978 New Orleans Data Base Design Workshop Report," Research Report RJ2554 (IBM Research Laboratories, San Jose, CA, July, 1979. Matthews, R.L. (1967), "A Computer Programming Approach to the Design of Accounting Systems," ABACUS, (December, 1967), pp. 133—1520 McCarthy, W.E. (1982), "The REA Accounting Model: A Generalized Framework for Accounting Systems in a Shared Data Environment," The Accounting Review, (July, 1981), pp. 7-13). (1980a), "Construction and Use of Integrated Accounting Systems with Entity-Relationship Mbdeling," in P.P. Chen, ed., Entity-Relationship Approach to Systems Analysis and Design, (North Holland Publishing Company, 1980), pp. 625- 637). (1980b), "Multidimensional and Disaggregate Accounting Systems: A Review of the 'Event' Accounting Literature," MAS Communication (July, 1981), pp. 7-13). (1979), "An Entity-Relationship View of Accounting Models," The Accounting Review, (October, 1979), pp. 667-686. McCrae, T.W. (1976), Computers and Accounting, (London: John Wiley and Sons Ltd., 1976). National Council of Physical Distribution Management (1978), Measuring Productivity in Physical Distribution - A $40 Billion Dollar Goldmine, (Chicago: National Council of Physical Distribution Management, 1978). National Council of Physical Distribution Management (1976), Annual Proceedings of the National Council of Physical Distribution Management, (St. Louis, Missouri, 1976). Pope, A.L. (1976), "The Concept and Cost Elements of Physical Distribution,” in Costing in Distribution, edited by M. Christopher and D. Ray (MCB Books, 1976), pp. 109-136). Ray, D. (1975), ”Distribution Costing - The Current State of the Art," International Journal of Physical Distribution, Vol. 6, NO. 2 (1975), pp. 73-1070 312 Rayburn, G. (1981), Marketing Costs - Accountants to the Rescue, Management Accounting, (January 1981), pp. 32-42. Rockart, J. (1979), "Chief Executives Define their own Data Needs," Harvard Business Review, (March-April, 1979), pp. 81-93. Sawdy, L.C. (1972), The Economics of Distribution, (John Wiley and Sons, 1972). Schiff, M. (1972a), "Physical Distribution: A. Cost Analysis," Management Accounting, (February, 1972), pp. 48-50. (1972b), Accounting and Control in Physical Distribution Management, (Chicago: National Council of Physical Distribution Management, Chicago, IL, 1972). Shirley, R.E. (1977), "Accounting Analysis of Distribution Activities--A Critique," International Journal of Physical Distribution, Vol. 7, No. 5, (1977), pp. 275-282. Sollenberger, H.M. (1971), Management Control of Information Systems Development, (New York: National Association of Accountants, 1971). Sorter, G. (1969), "An 'Events' Approach to Basic Accounting Theory, The Accounting Review, (January, 1969), pp. 12-19. Spaakman, G. (1981), "The Management Accountant and Productivity Improvement: Responsibilities and Techniques,” Cost and Management, (May-June, 1981) pp. 2-8. Stevenson, R. (1977), "Managing Physical Distribution," The CPA Journal, (May, 1977) pp. 74-79. Stewart, W.M. and J.E. MOrehouse (1978), "Improving Productivity in Physical Distribution: A $40 Billion Goldmine," in Proceedings NCPDM Annual Meeting, (Chicago, National Council of Physical Distribution Management, 1978), pp. 1-33. (1974), "Don't be Among the 99 44/100% of Companies Who Don't Know Their Distribution Costs," Handling and Shippin , (Presidential Issue, 1974), pp. 31-97. (1969), "P.D. Revisited," Proceedings of the NCPDM Fall Meeting,_ (Chicago, National Council of Physical Distribution Management, 1969). Taggart, Jr., W.M., and M.O. Tharp (1977), "A Survey of Information Requirements Analysis Techniques," Computing Surveys, (December, 1977), pp. 273-90. 313 Tavernier, G. (1975), "Controlling the Soaring Cost of Distribution," International Management, (August, 1975), pp. 11-16 0 Teichroew, T.J., and E.A. Hershey (1977), "PSL/PSA: A Computer Aided Technique for Structured Documentation and Analysis of Information Processing Systems," IEEE Transactions Software Engineering, (SE-3, l 1977), pp. 41-48. Teorey, T. and J. Fry (1982), Design of Database Structures, (Prentice-Hall, 1982). and J.P. Fry (1980), "The Logical Record Access Approach to Data Base Design," Computing Surveys, Vol. 12, No. 2, (June, 1980), pp. 179-211. Wait, D. (1980), "Productivity' Measurement: A. Management Accounting Challenge," (Management Accounting, May 1980), pp. 24-30 0 Wayman, W. (1972), "Harnessing the Corporate Accounting System for Physical Distribution Cost Information," Distribution System Costing: Concepts and Procedures (James R. Riley, Symposium on Business Logistics, April 1972), pp. 31-46. Willis, R. (1977), Physical Distribution Management, (New Jersey, Noyes Data Corporation, 1977). Yao, S.B., D.L. Navathe and J.L. Weldon (1978), "An Integrated Approach to Logical Database Design," in New York Symposium on Database Design, (Graduate School of Business Administration, New York, N.Y., May, 1978), pp. 1-14. .III‘ ‘ W ' I. - I, I4.