0...;n, AN EVALUATION OF THE INTERSTATE MANAGERIAL STUDY CLASSIFICATION OF KNOWLEDGE SITUATIONS Thosis for the Dogma of M. S. MICHIGAN STATE UNIVERSITY Curtis Franklin Lard I959 - i~r~4t~Afi~w Wl £5 rifle ARR mu 19m 2: om AN EVALUATION OF THE INTERSTATE MANAGERIAL STUDY CLASSIFICATION OF KNOWLEDGE SITUATIONS By Curtis Franklin Lard AN ABSTRACT Submitted to the College of Advanced Graduate Studies of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Economics Year 195 9 Approved Curtis Franklin Lard ABSTRACT The main purpose of this study is to evaluate the relevance of the "knowledge situation" concepts used in the Interstate managerial Study. The hypotheses tested by this investigation are: (l) farm managers en- counter and can recognize the knowledge situations in their own expe- riences; and (2) farmers' ability to recognize and give verified examples of the knowledge situations is related to their characteristics. The classification of knowledge situations under consideration is the one defined by G. L. Johnson and C. B. Haver. It is an extension of an earlier classification of Frank Knight's. The Interstate Managerial Survey provided the data for testing the hypotheses of this Study. A set of questions defining each of the know- ledge situations-pgsitive and negative risk agtigg, learnin , inaction, positive forced action and subjective certaintyh- and calling for cor- responding examples of the situation was. asked 1075 farmers in seven midwestern states. The questions were so constructed that the farmers responded with examples from their own managerial activities. In order for the answers to be considered complete, the farmers were asked to ex- plain what was done and what was involved in doing it. The complete or verified examples of the knowledge situations which farmers gave were used to test the hypotheses. The characteristics of these farmers were then related to their ability to recognize and give verified examples of the knowledge situations. The most important variables found to be related to ability to recognize and give verified examples of the knowledge situations were (in order): (1) education, Curtis Franklin Lard (2) thinking method used (induction and/or deduction), (3) attendance at agricultural meetings, (h) average gross farm income, (5) total debts, and (6) years farming experience. In general, farmers' ability to understand and give verified examples of the knowledge situations increases with; (a) higher education, (b) increases in the use of deductive reasoning, (0) increases in the number of agricultural meetings attended, (d) higher farm incomes, (e) increases in debts and (f) increases in years of farming experience. It was concluded that the classification of knowledge situations, as followed in the Interstate Managerial Survey, corresponds, to an important degree, with farm manager behavior. However, the classification was found to be incomplete. To further complete the classification, a new knowledge situation--involunta£y_learniEg--is added. Involuntary learning is defined as a situation wherein the manager does not voluntarily learn more since the cost of additional information equals or exceeds its value to him but, in which, some outside force makes it necessary to learn or for some learning to occur regardless of the will of the manager. The definition ofva new development (new technology) as followed in the IMS led to confusion of the knowledge situations. To help eliminate this confusion a new definition is presented. The definition of new technology, as advanced herein, is stated as follows: a technology will be considered "new" to an individual farmer until he makes either a positive or negative risk action decision toward adoption of the technology, after which, it is an old technology to him. AN EVALUATION OF THE INTERSTATE MANAGERIAL STUDY CLASSIFICATION OF KNOWLEDGE SITUATIONS By Curtis Franklin Lard A THESIS Submitted to the College of.Advanced Graduate Studies of Michigan State University of.Agriculture and.App1ied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Economics 1959 AC KNOE'IIEDGEI'EI‘H‘ S The author wishes to express his gratitude to the many people who have had a part in the development and completion of this thesis. The author sincerely appreciates the guidance and the constructive criticisms received from.Dr. Glenn L. Johnson during the course of this study. Special thanks are extended to Dr. Les V. Manderscheid for his helpful criticisms and assistance in preparing the manuscript for final typing. Thanks are also due to Dr. L. L. Boger and the Department of Agricultural Economics, for providing financial assistance during the program of graduate work and to fellow graduate students who have participated in any way in the development of this thesis. Finally, the author is indebted to his family for their encouragement and especially to his wife, Kathleen (Jimmie), for typing the earlier drafts of this manuscript. Any errors remaining herein are the sole responsibility of the author. CHAPTER I. II. III. IV. V. VI. TABLE OF CONTENTS INTRODUCTIONOOIQo.oOOCOOOOOOOOOOOOOOOOO00000000000 The Order Of Presentation...................... TIE TEEORETICAL SETTIIIG ELIND BACKGROUI‘ID. o o o o o o o o o o O The Role of Learning and Knowledge in the Management PTOCeSSoooooo00000000000000.0000. A Theory Of Knowledge.......................... Theoretical Development of the States of Knowledge................................... Defining the States Of Knowledge............... SOLTRCE OF DATA-coo.000000000000000000000000.0000... Interstate Managerial Survey................... The Survey.................................. The Sample.................oo...o........... Interviewing................................ COding ProcedureSooooo00000000000000.0000... A REPORT OF A PILOT STUDY.AND THE STATES OF ICNOI'IIIEDGE AS ST'U'DIES IN, E'ISoooooooooooogoaoooo. IX Pilot Study in Kentuckjr...OOOOOOOOOOOOOOOOCOC Definitions and Explanations of the Knowledge Concepts as StUdied in IMSooooooooooooooooo. The Control Variables.......................... THE RECOGNITION OF KNOULEDGE SITUATIONS........... Confusion of Knowledge Situations.............. The Importance of Each Knowledge Situation..... Summary.......................................o II‘TOPJ~LATIOII AND CONFUS IONS CONCERN ING TIE KNOW- MGE SITTfimTIONJOOOOOOOOOOOOOOOOOOOOOOOOOOO.... Types of Information by Knowledge Situation.... Confusion of Knowledge Situations by Type of Information InVOlvedOOOOOOOOOO00000000.000.0 smryOOOOOOOOOO0.0.0.000...OOOOOOOCOOOOOOOOOO PAGE 22 22 25 29 32 35 35 38 39 L12 hB TABLE OF COI‘ITEI‘ITS - Continued CHAPTER VII. ATTRIBUTES OF FARl-IIRS IN DEFEENT IS‘IO‘.JIEDGE SITUATIONS, SOURCES OF CONFUSION AND REIFORI’IUMTIONSooooococo-oecoooooooooooooooooooo Statement of Other Ex Ante and Ex Poste Hypotheses................................o. Characteristics of the Farmers Confusing Know- ledge SituationSooooooooooooooooo00000000000 Possible Explanations and Means of Eliminating Confusion of Knowledge Situations . . . . . . . . . . . The Definitions of the Knowledge Situations Involved.................................... The Definition of a New Knowledge Situation.... The Definition of a New Development (New TeChHOlOEY)0.000000000000000...-oucoooooooo The Distinction Between Alternative Approaches to Decision Making Under Uncertainty. . . . . . . . The Probabilistic Approach. . . . . . . . . . . . . . . . . . The Game Theoretic Approach. . . . . . . . . . . . . . . . . summaryOOOOOOOOOOOOOO00.0.0.0000.000.000.000... VIII. SUMI‘IARY, COINCLUSIOI‘IS AND DPIIICATIOI‘ISQQ0.00.0.0... SmaIyOOOOOOOOOOOOto...OOOOOOOOOOOOOOOOOOOOOO. The General Conclusions........................ Some POSSible ImplicationSoooooo000.000.00.000. Implications for Farm I~ianagement Research. . . Implications for Farm I~Ianagement Teaching. . . Implications for Extension Workers . . . . . . . . . . Implications for Farm Managers. . . . . . . . . . . . . . Implications for Agricultural Policy Formulation........o..................... AP13m\mL{OOOOOOOOOOOO0.0.0.0...O.00...OOOOOOOOOOOOOIOOOOOOOOO BIBI’l—OGPJLPIHOOOOOOOOOOOOOOOCOOOOOOOOOOOOOOOOOOOOOOOOOIOOOOOO PAGE A6 If? 50 53 5h 55 56 6O 61 62 6h 66 69 71 71 72 7h 7h 76 77 TABLE 1. 2. 3. 9. LIST OF TABLES PAGE Characteristics of the Sample of Eight Strata for the Interstate Managerial Survey oooooooooooooooooooooo l9 Farmers Responses to Questions Involving Understanding of and Encounters with Different Knowledge Situations.. 3h Number Farmers Confusing Each Knowledge Situation with the Other Knowledge Situations .................... 36 Number of Times Components of Each of the Six Najor Types of Information were mentioned by Selected Groups of Midwestern Farmers in l9Sh .................. hO Number of Examples the 1075 Farmers Gave by Type of Information in.Each Knowledge Situation and Pur- pose Needed IOOOOOOOOOCOOOOCOO0...O.................... hl Percentages of Farmers' Examples Given by Knowledge Situations within.Each Information Category ........... AZ The Confused.Examples of Knowledge Situations by Type Of Informtion In'VOIVed o00000000000000.0000...coo Ml A Comparison of the Attributes of Farmers' Ability and Inability to Recognize and Give Verified Examples of One Knowledge Situation and that of Other Knowledge Sitllations OOOOOOOOOOOOOOOOO00.0.0.0...0.00.00.00.00... 51 A Comparison of the Characteristics of the Farmers Confusing Knowledge Situations with Those Who Gave uncorl'fused verified mples 0.000000000000000000000... 52 LIST OF FIGURES FIGURE 1. 2. 3. h. 5. Cost and Value Functions for Knowledge The Risk Action Knowledge Situation . The Learning Knowledge Situation . . . The Inaction Knowledge Situation . . . The Forced Action Knowledge Situation PAGE 26 27 27 28 28 CHAPTER I INTRODUCTION This study is an integral part of a group of investigations, known as the Interstate Managerial Study (hereafter referred to as IMS), con- cerning the managerial concepts and principles employed by farm.managers in operating their businesses. The specific objectives of this inquiry are (l) to test the general hypothesis that farm.managers recognize the different degrees of knowledge when they are explained to them, (2) to see if the managers can give verified examples of having encountered these knowledge situations, (3) to make and test certain hypotheses about the attributes of the farmers who recognize and give verified examples of the knowledge situations and the types of information in- volved, and (h) to state the general implications of the study. The degrees of knowledge under consideration are those defined by two midp 'western agricultural economists1 who have extended an earlier classi- fication of Frank Knight's.2 The study addresses itself both to formulating concepts about "states of knowledge” and to empirical tests of hypotheses involving these concepts. This part of the IMS tests the validity of these con- cepts and attempts the resultant necessary reformmlations. Knowledge is the end product of learning which is a necessary task 1G. L. Johnson and C. B. Raver, Decision Makin Princi les in Farm.Management, Kentucky Bulletin 593 (Lexington: UEIversIty of Kentucky, 19 3 . 2Frank H . Knight, Risk. Uncertainty and Profit, Houghton-Mifflin, New York, 19210 -2- of most people operating a going concern, such as a farm. The knowledge a manager possesses at any given time may vary from.none to perfect and his behavior is very likely to vary with the amount or degree of knows ledge possessed. It is a part of this study to offer explanations and describe the characteristics and behavior of farm managers possessing different spec- ified amounts or degrees of knowledge. Both the cost of obtaining and the value of possessing knowledge are subjective. Therefore, in studying the knowledge a manager possesses, we must use subjective measurements. As the questions used in.this study from the IMS, were most frequently stated so that a manager told or responded about his own personal expe- riences,3 the data are often behavioristic. The question is frequently asked by economists, "What use can be made of information on the degrees of knowledge held by farm managers?. This study should provide a better understanding of how farm managers think when managing their farms. It is believed that such information would help direct or improve a) research in farm.management, b) the type of information made available to farm.managers, c) the extension aid offered to farmers, d) the education of the farm management student and e) policy recommendations for agriculture. The farm firm is both a decisionpmaking producer and a decision- making consumer. The IMS takes the interdependency of the farm.firm.and the household into consideration. The Order of Presentation The basic plan of this study may be described as follows. 3The list of questions concerning the knowledge situations, as studied in the IMS, is present in the Appendix. -3- In Chapter II, the theoretical setting and background for the states of knowledge are developed. Included in this chapter are (1) the role of the learning function and knowledge in the managerial processes, (2) a theory of knowledge, (3) the theoretical development of the states of knowledge over the period from the classical economists to present and (h) definitions of the "states of knowledge.“h The origin and objectives of INS will be given in Chapter III. Attention will be directed toward the survey, the sample, the interviewing and the coding procedures. 'While the general overall characteristics of the IMB will be described, major attention will be given to that part concerning knowledge situations. In Chapter IV, the primary purpose is to set the stage for this investigation. .A summary is presented of the findings of a pilot study in Kentucky".5 This pilot study was a forerunner to the IMS wherein certain questions and concepts concerning the managerial function were tested for appropriateness and relevance. Also in Chapter IV, the knowledge concepts are defined and described as they have been studied in no.5 In the last section of this chapter, the independent variables are described and discussed. The three chapters that follow (V, VI, and VII) present the empir- ical evidence and relations tested in this study; Chapter V is COD! cerned with presenting the evidence of the recognition of knowledge hJohnson, op. cit. p. 11f. 5G. L. Johnson, Manggerial Concepts for.A riculturalists, Kentucky Agricultural Experiment Station, Bulletin 619, 1955, p. 33?. 6Great Plains Council. Proceedings of'Research Conference on.Risk and Uncertainty; Great Plains ouncil Publication no. II. Fargo, NoFth Dakota: North.Dakota.Agricultural College, 1953. -h- situations and the testing of the general hypothesis that farmers encoun- ter and can recognize the different degrees of knowledge. In this chapter, knowledge situations confused and possible explanations for such confusions are presented. Also, the importance of each knowledge situation is stressed. In Chapter VI, the types of information used in solving management problems are presented, along with, the types of information involved when managers confuse knowledge situations. In.Chapter VII, which.is the real core of this analysis,thore are stated other ex ante and ex pgste hypotheses which have been verified or rejected with.the empirical evidence from.IMB. Also, some possible means are offered for eliminating the confusion of knowledge situations. Among these possible means the following are presented: (1) a new definition of new technology, and (2) a distinction between the game theoretic and the probablistic approach to decision making. The general conclusions and implications made from.this study are given in Chapter VIII. Some possible implications will be given for further research, farm management teaching, extension work, farm managers, and public policy formulation. These implications are based on observations and insights gained in this study. CHAPTER II THE THEORETICAL SETTING AND BACKGROUND In developing the theoretical setting for the analysis, it is necessary to examine the learning process to see why it is necessary in management. This examination will be made by reviewing the past work on the degrees of knowledge. Therefore, such examination will make obvious the importance of an empirical investigation of the know- ledge situations. The Role of Learnigg_Process and Knowledgg in the Managerial Process The management process is divided into five distinct steps.7 Assuming the manager has defined the problem.and determined what should be observed, the five tasks of management are as follows: (1) Observation (2) Analysis (3) Decision making (h) Action (5) Acceptance of responsibility Of the five steps, the first three involve learning. Knowledge is important in the managerial process. The role of learning in the managerial process is the acquisition of information and knowledge. The learning process is one means whereby the manager improves his decisions. The need for learning exists because of the gap between predictions 7Johnson and Raver, op. cit. page 8. -6- and realizations. The more correctly the manager can predict the future, the better manager he is and the higher rewards (profits) he will re- ceive. T.'W. Schultz defines the entrepreneur as one who must make two interrelated decisions, what to do and how to do it.8 In the same article, Schultz states, "Little has been done to reduce the divergence between expectations and realizations. The gap between expectations and realizations is a positive measure of what is probably the most important source of inefficiency and waste in present day (1939) farming. ..... mistakes are usually traceable to the fact that expectations were faulty." If such statements are accepted, then the learning process and the degrees of knowledge become the basis for improving the efficiency of managers and reducing economic waste. Because knowledge is so often imperfect, it is necessary that we investigate the theory of knowledge to develop a fuller understanding of the subject matter, herein presented. In this study, no attempt is made to give a complete description of the theory of knowledge (which is not well developed anyway). Enough should be described, however, to indicate the general nature of the learning process as it is related to managerial decision making. It should also stress the degrees or correctness of knowledge in the process of developing a fact or "true" knowledge. In a dynamic economy -- where factors affecting the firm.are changing continuously -- a manager finds that his knowledge is often imperfect and incomplete. This makes it necessary for the manager to learn. Since learning requires time, effort, and money, the manager attaches a 8Schultz, T. W., "Theory of the Firm and Farm Management Research," Journal of Farm Economics, Vol. XXI (1939). -7- disutility to the process of acquiring additional information. If we assume the law of diminishing returns applies to the learning process, then it would yield an increasing marginal cost curve and a diminishing marginal utility curve for knowledge as the product of the learning process. Then the marginal disutility of getting information can be equated with the marginal utility (or worth) of such knowledge to max- imize income or total satisfaction. A Theory'of Knowledge Science can be regarded as storehouse of organized, consistent knowledge and the scientific method as a process whereby we find truth or knowledge. ‘Whitney says, in the Elements of Research, '..... after every pause for analysis, integration, and deductive verification, a leap more or less in the dark must be made, if any conclusion at all, however tentative, is to be reached." Thus, knowledge may not be exact but as Dewey says Iscience is simply the most authentic knowledge of nature, man, and society that is possible at any given time by means of the methods and techniques then and there available.'9 .As knowledge is ordinarily only tentative and imperfect, a manager in an ever-changing dynamic economy must continually employ himself in the learning process. A manager may have imperfect knowledge which can be overcome to a certain extent by learning. People can employ two methods of learning -- inductive and deductive. Deductive learning (or reasoning) is conceived to be concerned with the conditions under which particular or instanial propositions are inferable from universal premises. Inductive reasoning, on the other hand is conceived as dealing with those inferences which 9A quote from Elements of Research by Whitney. -8- enable us to derive universal conclusions from particular or instantial premises. Deduction, as contrasted with induction, is distinguished by' the fact that the conclusion is certain and necessary if the premises are.. In general, conclusions reached by induction are probable only. Deduction proceeds from general principles to other general principles or to particulars; induction seeks to establish general principles or laws by examination of particular cases. Deduction is analytic; induction, synthetic. A set of more general definitions may be stated as follows: (1) deductive learning - a method of scientific reasoning by which from general principles concrete applications or consequences are deduced and (2) inductive learning - act or process of reasoning from a part to a “11019 010 Knowledge as acquired by scientists is based upon rigorous demonstration and consistency, while the knowledge which an individual agg_layman considers adequate may be less rigorous and of a personal subjective nature. Knight says, "The ordinary decisions of life are made on the basis of 'estimates' of a crude and superficial character. In general the future situation in relation to which we act depends upon the be- havior of an indefinitely large number of objects, and is influp enced by so many factors that no real effort is made to take account of them all, much less estimate and summate their sep- arate significance. ..........Propheqy seems to be a good deal like memory itself, on'which it is based. So when we try to decide what to expect in a certain situation, and how to be- have ourselves accordingly, we are likely to do a lot of irrelevant mental rambling, and the first thing we know we find that we made up our minds, that our course of action is settled.'11 1°Cohen, M. R., and Nagel, E. An Introduction to Lo ic and Scientific Method. New York: Harcourt, Brace and Co., 1955. 11Knight, op. cit. -9- Marshall12 has remarked that the business manager's decisions are guided by'”trained instinct" rather than knowledge. Whitney13 remarked that the normal human mind often acts in terms of prOblem situations without thought, i.e. tradition or habit are sub- stituted. Once a person has experienced an event, he may need very little extra information to decide when he is faced with the similar problem.again. Often times it may be advantageous to act according to custom or habit, since cost of acquiring the extra information may exceed its value. It is evident from the brief discussion of the nature of knowledge above, that knowledge may be of varying degrees. A classification of degrees of knowledge for use in studying management will be more useful if it considers the personal, subjective value and costs of knowledge as well as the more objective measures. Theoretical Development of the States of Knowledgg It is now appropriate to classify the different degrees of knowb ledge. Knight distinguished three different degrees of knowledge. They are perfect knowledge or certainty, risk and uncertainty. He defined perfect knowledgg or certainty as a situation in which a manager has no risk-bearing and learning task to perform. This degree of knowledge is lessentially'that assumed by the classical static economic theorists. Knight defined EEEE as where probabilities of making errors of perception and inference are known, thus permitting the risk-bearing function to be carried out. However, since the probabilities of errors are known, the costs of bearing the risks can be computed and incorporated into an 12Marshall,.A. Principles of‘Economics. London: Macmillian Co., Ltd., 1936. DWxitney, op. cit. -1o- insurance scheme, thereby eliminating this type of risk bearing as a necessary managerial task. Uncertaintyjis not susceptible to meas- urement and cannot be eliminated. It is this uncertainty which gives rise to the management function and profits. Hart pointed out that Knight's distinction.between risk and uncertainty was somewhat incom- plete. Albert G. Hart1h recognized the lack of clarity in the distinction between Knight's risk and uncertainty. Hart defined EEEE to denote the holding of anticipations which are not "single valued“ but constitute a probability distribution havingkzmwn parameters. He defined uncertaingy to denote the holding of anticipations under which the parameters of the probability distribution are themselves not single valued. He argued that if known probabilities exist and the entreprenuer has the pos- sibility of deferring decisions (with or without special costs) the manager can still be in an uncertainty situation - because he is willing to forego decision to learn more or collect more evidence to improve the accuracy of his prediction. This is to say, the manager may know the probabilities of error and be able to compute risk-bearing costs but feel that the passage of time will permit him to learn more about the event at less cost than the value of such knowledge. Thus, he may act as if the situation were an uncertainty situation. . The important contribution of Hart's analysis was that he antic- ipated the close correspondence between the process of improving esti- mates with the passage of time and the principles of sequential analysis. Hart said, "Unless the event in question is imminent, the future must be 11a. G. Hart, "Risk, Uncertainty, and the Unprofitability of Compounding Probabilities," Reprinted in.§£udies in mathematical Economics and Econometrics, l9h2, Pages llO—lIB. -11- expected to bring in more relevant evidence. Possibly new evidence will change our outlook and give our estimates a radically dif- ferent expectation value. more probable new evidence will con- firm our impressions and leave the expection value substantially unchanged."15 This is to say, the manager tries to preserve flexibility or con- tinue learning in his planning process and in his firm, depending upon the cost of the new evidence and of delay in reaching a decision. Abram Wald,16 a statistician, re-examined the formal theory of statistical decisionpmaking in the late thirties and early forties. In single sample analysis, statisticans set up certain standards of accuracy and compute the sample size necessary to reach such accuracy, after which, they make the observations and make their terminal decision. weld devised a system of statistical decision making whereby a standard of accuracy is set up first and evidence is then gathered and analyzed simultaneously. This is called sequential decision --- a series of decisions or a chain. The word sequential deals with the situation in which evidence is collected in little units (observations) one at a time and the information at each stage is used to make the choice among three decisions: (1) enough information is available to accept one alternative, (2) enough information is on hand to accept the other alternative or (3) a decision to take still another observation. If the decision is to take on additional observation, the learning process con- tinues. (This formulation made Hart specific.) A person using the sequential process needs the capacity to sort out the incoming information (which theory specifies) to determine what 151bid., pp. 110-118. 16Abram Wald, Statistical Decision Functions, John Wiley and Sons, New York, 1950. -12- is incomplete, unreliable, biased or irrelevant. The accuracy of data is important since the decision is only as good as the data used to fuel the statistical decision-making system. In decision making, the decision maker relies on both objective and subjective measurements. Subjectivity is involved because the accuracy of information is often based on the personal evaluation of the decision maker. The development of sequential statistical analysis provided the basis for dividing imperfect knowledge into subjective risk situations and three sub-categories of subjective uncertainty. The distinction be- tween subjective risk and subjective uncertainty depends upon the stan- dard of accuracy required by the manager or person desiring the decision. The three subjective uncertainty situations are (l) the situation wherein learning is continued, (2) the situation where learning is discontinued because its cost exceeds its value to the analyst or no action is taken because not enough information is available for a decision, and (3) the situation in which a manager (or person desiring the decision) is forced to act by outside circumstances even though more learning would be worth- while if time permitted.17 G. L. Johnson18 viewed Knight's thinking as being incomplete as follows, _ "First, he distinguished between risk and uncertainty on the unrealistic and objective basis of whether or not it was possible to compute probabilities of errors rather than on the subjective, but more realistic basis, of whether or not the amount of infor- mation on hand was considered adequate for action. Second, he did not break his uncertainty category down into sub-categories distinp quishing between situations in which managers try to learn, do not try to learn, and are prevented from learning." In the late forties, G. L. Johnson19 defined four degrees of knowledge 17These situations are more clearly and distinctly defined on page 134. 18Johnson, Kentucky Bulletin 619, op. cit. 19Johnson, G. L., unpublished doctoral thesis, "Allocative Efficiency of.Agricultural Prices," University of Chicago, l9h9. -13- possessed by managers while in the decision making process. These four knowledge situations were (1) certainty'where the manager knows the future with certainty and has no risk-bearing function, (2)‘£i§§ wherein a manager knows that a future event will fall within a known probability distribution, (3) uncertainty wherein a manager may know a future event will fall within a series of probability distributions to each of which is attached a likelihood, and (h) non-certaingy;wherein the manager may’ know nothing about a future event. Since Johnson wrote his doctoral dissertation, he has reformulated and reclassified the knowledge situation held by managers. For a rede- fining and reformulating of the degrees of knowledge see the discussion belovezo Defining the States of Knowledge The degrees of knowledge were more clearly and distinctly formulated by’Johnson,21 when he was a member of the.Agricultural Economics staff at the University of Kentucky, in 1950. He did some reformulating of the knowledge situations as they were presented in his Ph.D. dissertation, along with, added information which he gathered by doing some case studies. These few case studies in two Kentucky counties were based on responses to probing, open-ended type questions used when he interviewed a few farmers to see how near these concepts coincided with their be- havior. .As a result of the extra information received from.the case studies and other research work in collaboration with L. A. Bradford (on some occasions) and C. B. Haver (on other occasions), Johnson presented and defined five knowledge situations. These five knowledge situations 20Also see Kentucky Bulletin 593 and North Dakota Bulletin too. 21L..A. Bradford and Glenn L. Johnson, Farm.Management Analysis, John Wiley and Sons, New York, 1953. ~1h- in which farm managers find themselves are as follows: (l) subjective certainty, and (2) subjective uncertainty including (a) risk (b) learning (c) inaction and (d) forced action. These were defined as follows: (1) subjective £355 is where information, though known to be imperfect, is considered to be adequate for decision and the cost of more information equals its value, (2) learning is where the manager incurs additional cost to get additional information because the knowledge present is con- sidered inadequate for decision and the value of additional information exceeds the cost of acquiring it, (3) inaction is where present knowledge is inadequate for decision and the manager has no ability or desire to learn more since the cost of information is higher than the value, (u) forced action is a situation wherein a time element is involved and some outside element forces the decision before the manager is able, willing and ready to bear the consequences. The manager regards present knowledge as inadequate for decision but the "time element" prevents further learning. Subjective certainty is defined as a situation where a manager regards present information as perfect or so nearly perfect that he can make the decision and ignore probabilities of errors. After the case-study, mentioned above, a.more detailed pilot study was conducted in Kentucky. From this pilot study, the knowledge situa- tions were redefined by Johnson and Haver22 as presented in Chapter IV. Also in Chapter IV, a graphic representation and the definitions of the knowledge situations will be given as they have been used in carrying out this study. 22Johnson and Haver, op. cit. p. ll ff. CHAPTER III SOURCE OF DATA Interstate Managerial Survey The Interstate Managerial Study, hereafter referred to as D6, is based upon the ideas, principles, and concepts of management as they were stated in the bulletin, Decision Making Principles in Farm Manage- E293. by Johnson and Haver.23 The min contribution of the bulletin, mentioned above, is a functional-situational framework within which farm management may be viewed. Assuming the problem has been defined, the five functions or tasks that management performs are: (1) observation, (2) analysis, (3) decision making, (14) action, and (S) acceptance of responsibility. In carrying out these tasks (especially the first three), the managers encounter different degrees of knowledge which are (l) subjective uncertainty, (2) subjective certainty. These different degrees of knowledge were believed to be possessed with reference to the know- ledge which the manager had about a type of information. The types of information used by farmers in decision making were classified as (1) price, (2) production methods, (3) new development, (1;) human, (5) institutional, and (6) home technology. This bulletin formed the basis for the DB. In August 1953, a Risk and Uncertainty Conference was held by mid- western agricultural economists at Bozeman, Montana. At this conference, 23Johnson and Haver, op. cit. -lb- an interstate survey was decided upon as a.means of obtaining data to test the concepts and principles set forth in Johnson's and Haver's bulletin e The Survey?!4 Cooperative relationships were established which ineluded agricultural econaaics personnel interested in the development of managerial concepts and principles from the Agricultural Experiment Stations of: (l) Ohio State UniverSlty, (2) University of Kentucky, (3) Purdue University, (a) Michigan State University, (5) North Dakota State Agricultural College, (6) Iowa State College, and (7) Kansas State College. The services of the Farm Foundation and of the Risk and Uncertainty Sub-committee of the North.Central Farm Management Research Committee were utilized in estab- lishing these cooperative relationships. These institutions cooperated in setting up and running the survey. The survey was conducted in seven states during the summer of l9bh. A total of lU75 schedules were completed on farmers in the lOllowing states: (l) Kentucky, (2) Indiana, (3) North Dakota, (h) Iowa, (5) Kansas, (o) Ohio, and (7) Michigan. Michigan State University, as originator and a primary sponsor of the survey, arranged for and con- tributed the serVices of a survey expert for use (a) in constructing and pro-testing the survey schedule and (b) in training interviewers. The Sample The area and units to be sampled were delineated by the North Central For a more detailed description of the survey, see the Journal of Farm Economics Proceedings No. b, (uecember, lybb), pages lU97-llu9- -17- Regional Committee. The area consisted of eight geographical regions containing contiguous groups of whole or part counties located within the seven states (mentioned above). The units interviewed consisted of non-urban commercial farms (census definition) with gross farm income of $2500 or more and which 'were managed by'a single household unit. Commercial farms characterized by livestock share leases, father-son arrangements where both had a separate family and household, and regular business partnerships between two unrelated individuals were not eligible for interview. The survey was conducted by trained interviewers who were instructed to complete a specified number of schedules in each of the eight areas by the members of the Committee. It was decided that a stratified random sample using area sampling units would be the appropriate design. Each of the eight areas was a stratum and each stratum was subdivided into area sampling units which contained on the average two eligible farms (in the case of Kentucky it was decided to use area sampling units which contained on the average three eligible farms). The decision to use the above sampling unit sizes in terms of eligible number of farms was based on considerations of cost and ease of field operation. The actual sample drawing was completed using the 1950 Census of Agriculture and the 19147 Revised Master Sample Materials. It was first necessary to determine the numbers of eligible farms present in each whole or part county. This number was obtained for each county by taking the 1950 number of commercial farms with gross incomes of’82500 or more; subtracting from this the 1950 number of livestock share leases and finally'reducing this number by 20 percent in order to -18.. adjust for partnerships, father-son arrangements, and changes in the number of farms since 1950. The 20 percent reduction factor was arrived at through the experience of the staff of the Sample Survey Group of the Statistical Laboratory at Iowa State College. After having determined the total number of eligible farms in each county, the total number of area sample units Within that county was determined, and then the Master Sample Materials were used in subdividing the county into area sampling units of the desired size. A simple random sample of the desired number of area segments was drawn fromieach stratum (with the exception of stratum 8) and these were numbered and indicated on % inch scale county highway maps. The area segments within each stratum were numbered serially in the order with which they were drawn and the number assigned to segments on the maps consists of the stratum.number followed by the area segment number. In the case of stratum.8 (Kansas), the number of interviews to be obtained was allocated to the individual counties by the Kansas representative on the North Central Regional Committee. A simple random sample of the desired number of area segments was drawn from each county in the stratum and these were indicated on county highway maps in the same manner as the other strata. The characteristics of the sample of the eight strata are presented in Table l. Interviewing In June 1954, an interviewer's school was held at the Purdue University for one week. The school was instructed by Joel Smith of the Sociology Department of Michigan State University. G. L. Johnson and.A. N. Halter assisted in instructing the interviewers. The interviewers attending -19- TABLE I CHARACTERISTICS OF THE SAMPLE OF EIGHT STRATA FOR THE INTERSTATE MANAGERIAL SURVEY Estimated Estimated Expected Actual Number of Eligible Number of Number of State Eligible Farms per Interviews Interviews Farms Sampling Taken Unit Kentucky 1,790 3 150 12h Ohio 23,599 2 200 137 Indiana 15,769 2 200 189 Michigan 37,516 2 221; 199 Michigan 39h 2 30 30 North Dakota 9,301 2 150 129 Iowa 23,6h9 2 11:0 120 Kansas 6,985 2 206 1h? represented each of the states involved in the study. The purposes of the school were (1) to acquaint the interviewers with the study, the survey and the schedule; (2) to instruct the interviewers in the proper techniques to use in interviewing; and (3) to supervise some practical exercises in interviewing under field conditions. The interviewing was done in the summer and fall of l95h. .A total of twentybthree interviewers in the seven states contacted the eligible farms. The interviewers were given specific instruction of how to locate each prospective interviewee or farm. He was also given instructions of how to use each of the six schedules. These six schedules were rotated in sequence and each one contained the questions to test the specific -20- hypotheses of the study but not every schedule contained the same listing of questions.25 'When 10 to 20 interviews were completed by an interviewer, the com- pleted schedules were sent to Joel Smith for review. He examined the completed schedules for uniformity in interviewing and completeness. Coding Procedures The personnel at Michigan State University constructed a code which made it possible to transfer the data from the schedules to IBM punch cards. The process of coding consisted of four stages: (1) the con» struction of a preliminary code, (2) the revision and testing of the code, (3) the actual coding, and (h) the checking. The preliminary code was constructed by Joel Smith, G. L. Johnson, and A. N. Halter. They took a large number of responses to a specific question and defined categories into which each answer would fit. The answers were then assigned numbers. 4After the preliminary code was completed, it was presented to the Risk and Uncertainty Sub-Committee. Then.the code was revised in accordance with what the sub-committee recommended. The revision was tested for reliability. This test was done as follows: (1) two per- sons would code 15 or 20 actual questionnaires randomly selected from the seven states; (2) the code numbers assigned by the testers for each item were compared for agreement; (3) when the code numbers did not agree, appropriate adjustments were made until the coding became consistent. In the third stage, the actual coding of the schedules carried out by clerks under the supervision of Joel Smith and A. N. Halter. The 25The questions concerning knowledge situations were on every schedule. -21- clerks carried out the coding by the use of the code sheet. The coding was then checked by a second person. After all the answers had been assigned numbers by use of the code sheet, the tabulating department of’Michigan State University punched the code numbers into IBM cards. The data on each schedule required a total of hBO columns on six IBM cards. The punched cards were again checked for interrelated punches between the columns for each question. After initial marginal tabulations were run these checks were repeated on punch and column totals. CHAPTER IV A REPORT OF A PILOT STUDY'AND THE STATES OF KNOWLEDGE AS STUDIED IN IMS As has been discussed in Chapter II, the knowledge situation pos- sessed by'a manager depends upon his personal subjective evaluation. The previous discussion has provided a basis for sub-categorizing the uncertainty or imperfect knowledge situation into more meaningful, realistic terms. In the succeeding sections of this chapter, a review of the reSults of an empirical test of the knowledge concepts is pre- sented. Also, the definitions of the different degrees of knowledge (as studied in IMS) which the manager possesses are given in more detail. A Pilot Study in Kentucgy26 A pilot study, concerning the management concepts and principles, was conducted in Montgomery County, Kentucky in 1951. Thirtyrone farmers were asked questions concerning the management problems which they faced; the information they used in solving problems; the impor- tance of knowledge situations (risk, learning, inaction and forced action); the relative importance of inductive and deductive thought processes; whether or not they employ the flexibility principle; and the importance of strategic operations in their managerial activities. The primary objectives of this study were (1) the sorting of existing managerial theory and principles for relevance, (2) the studying and understanding of managerial activity. The study was designed to serve as a basis for formulating subsequent more adequate surveys, in 26Johnson, "Managerial Concepts for Agriculturists," op. cit. p. 36f. -23- particular the IMS. Most of the farms surveyed were over 60 acres in size, and were located on soils of mixed limestone and shale origins. The farms had fairly large burley tobacco bases and the farmers experienced moderate incomes. In general, the area was well served with roads, schools, markets, electricity, and telephones. Since this study was of an explor- atory nature, it was deemed unwise to attempt a greater geographic sur- vey. However, despite its geographic limitations, the survey was sur- prisingly conclusive in some instances and highly suggestive in other instances. From the pilot study, the following conclusions were reached: (1) the learning situation was empirically important; (2) farmers do take steps to prolong the learning situation, i.e., to keep the business flexible in order to gain from.what can be learned; (3) farmers have clearcut ideas about the nature and extent of the costs and values of flexibility; and (h) farmers weigh the costs of learning against the value or usefulness of what can be learned. In summary, the pilot study gave the following results: (1) the farmers indicated about a fifty-fifty split on the use of the inductive and deductive thought processes; (2) all five types of information (price, production, innovations, human, and institutional) proved to be important but the listing appeared incomplete; (3) subjective uncertainty situations were important including risk and forced action, though forced negative actions were confused with negative risk actions and with inaction due to lack of knowledge while the learning situation appeared to be particularly important, (b) all farmers indicated that they employed inductive reasoning and all farmers indicated that they employed deductive reasoning, and (5) all farmers indicated they employed personal strategies. -2h- While the above study was being completed, G. L. Johnson and C. B. Haver27 collaborated and came up with the following classification of knowledge situations; positive risk action, negative risk action, learning situation, inaction situation, certainty and positive forced action. These are defined as follows: (1) positive risk action - where a manager is not completely sure (he has a probability distribution of the outcomes) of the outcome, but is willing to take the consequences of acting and being wrong; (2) negative risk action - wherein a manager decides not to act even though he runs a risk of being wrong in not acting, he is willing to take the consequences of being wrong, by not acting; (3) learning_situation - wherein a manager postpones a decision to act or not to act until he can get additional information; (h) inaction - a situation wherein the manager does not have enough information to act but the value of additional infor- mation is not worth.the cost and effort of learning it; (5) certainty;- a situation wherein the farmer acts as if he were certain of the outcome and does not worry about being wrong; and (6) positive forced action - a situation wherein the manager is forced by circumstances to make a decision when he regards the information he has as inadequate for decision. He is forced to act before he is ready, willing and able.to bear the consequences. The sub-dividing of risk action into positive and negative becomes necessary because of the different consequences associated with the difb ferent actions. In deciding to act either positively or negatively, the 28 manager faces the possibility of making two types of errors, when choosing between two alternatives. The Type-I error is made when a hypothesis is 27G. L. Johnson and C. B. Haver, Decision.Makipg_Principles, in Farm Management, Ky. Bull. 593, 1953. 28J. Neyman and E. Pearson's work on this subject is summarized by P. Hoel, Introduction to Mathematical Statistics, John'Wiley and Sons, New Iork,lI9E7, pp. 202-206. -25- rejected when, in fact, it was true or best. The Type-II error is made when the alternative hypothesis is accepted when, in fact, it was false or worst. In decision making the manager Specifies the accuracy desired and the probability of making each of the errors. The different con— sequences involved, if either error is made, help determine whether the action taken Willee positive or negative. The knowledge situations, as defined by Johnson and Haver, are essentially the classification studied in IMS and used in this thesis. The questions used to study the different knowledge situations are pre- sented in the Appendix. Knowledge situations, as studied in IMS, are more vividly eXplained and described in the discussion which immediately follows. Definitions and Explanations 29 of the Knowledge Concepts as Studied in IMS In defining and explaining the knowledge situations held by managers, marginal analysis is used. Since a role of management is to narrow the gap between business eXpectations and realizations and this gap exists be- cause of imperfect knowledge, we can explain the nature of returns to managers in terms of their capacity to form correct judgments. Managers learn to improve knowledge and reduce risk at a cost. This cost involves money, time and effort. From the above discussion and the results of the "Pilot Study," the manager can be regarded as placing a marginal value on additional infor- mation. and incurring a marginal cost on the cost of acquiring such” information. Since the value placed on additional information is sub- jective and the costs involve something other than money cost, we can re- place marginal value with marginal utility and marginal cost with marginal "Learning Processes the Individual.Approach," in the 29G. L. Johnson, 1 g f se m...an lgricu tural Experiment Station, Bulletin too, 1955. -26- disutility. ‘We can assume the law of diminishing returns applies to learning, because statistical formulas show diminishing returns (in terms at accuracy) to size of sample and general experiences seem to indicate that there are diminishing returns in deductive thinking. In the process of acquiring information, there is the possibility that the cost of infor- mation increases at an increasing rate since as more and more learning is acquired it becomes more difficult (in terms of effort, time and money) to get an extra unit of information. It follows then that the cost of acquiring additional information yields an increasing marginal cost (or disutility) curve. The law of diminishing utility yields a diminishing marginal utility'curve for knowledge, the product of the learning process. These two curves, marginal utility and marginal cost, can be placed on the same diagram as shown in Figure 1. The value of additional information (MU curve) decreases atan increasing rate and the cost (MC curve) of additional information increases at an increasing rate. Personal value of 1 knowledge or cost of acquiring MC (disutility) knowledge MU (utility) fir Knowledge Figure 1. Cost and Value Functions for Knowledge. -27- The risk action situation is defined to exist where present knowledge, though known to be imperfect, is regarded as adequate for decision and the cost of additional knowledge (or information) equals its value. The risk situation is diagramed in Figure 2. The learning situation is defined as a case where present knowledge is inadequate for action in the sense that the manager is subjectively unwilling to decide and more knowledge can be acquired at a personal subjective cost lower than its value. Thus, he engages in further learning before deciding. This situation is diagramed in Figure 3. Here the flexibility principle is involved, i.e. the firm delays decision and keeps its business flexible so it can do more learning. It is often possible for a manager to move from this situation to a risk action situation. Personal Personal value or value or costs costs MC -a U) -a 0 3’0 853 E3 E's fig m a: n he 13% (DO) 03 n a: m 53 a: m o A: 3: Knowledge Knowledge Figure 2. The Risk Action Situation. Figure 3. The Learning Situation. The inaction situation is defined as a case where present knowledge is inadequate for action (which confirms the decision not to take positive action) and the cost of more knowledge exceeds its value - hence, there is no action and no learning. This situation is diagramed in.Figure h. -2 8.. The forced action situation is defined as a situation wherein present knowledge is inadequate for action but some outside force makes a decision (positive or negative action) necessary before more learning can take place. Decision and action are forced before the manager is ready, willing and able to act. This situation is diagramed in Figure 5. Personal value or costs Present knowl ge inadequate Personal value or costs 7M0 U) H O G.) "O L. 1 Knowledge Figure h. The Inaction Situation. MU : O «4 (0 od 0 as '8 u as Q) ’5 .p MC 88' g 'o s33 Knowledge Figure 5. The Forced.Action Situation. The certainty situation is defined as a case where present knowledge is adequate for decision, and the value of additional information is zero. The manager regards it as perfect or so nearly perfect that he can ignore probabilities of errors. No attempt has been made to diagram this knowh ledge situation. Because of the importance of negative risk action, it has been separated from positive risk action in this study. Negative risk action and the five situations described above are the ones which have been examined empirically in the IMS. -29- The Control Variables In testing the knowledge concepts empirically, certain variables were held constant or controlled. The "control" variables included: (1) education, (2) additional training, (3) past participation in h-H and FFA, (h) years farming experience, (5) age, (6) average gross income, (7) net worth, (8) total assets, (9) total debts, (10) acres in crOpland and rotation pasture, (11) thinking method used in arriving at conclusions, (12) most natural thinking process used by the farmer, and (13) the respondent's attendance at organizational meetings. In setting up the study of the knowledge concepts, these variables were conceived to have certain relationships with the ability of farmers to recognize and give verified examples of the different degrees of knowledge they had en- countered in their Own experience. It was hypothesized that farmers with more formal education would be more able to recognize and give verified examples of the different knowb ' ledge situations. The breakdown of the farmers by grades of formal education completed was as follows: (1) l to 7 grades -- 130 farmers; (2) 8 grades -- 355; (3) 9 to 11 grades -- 197 farmers; (h) 12 grades -- 286; and (5) 13 grades and over -- 102 farmers. It was hypothesized that farmers with more additional training could more easily recognize and give verified examples of the different knowledge situations. The additional training included such things as (1) G. I. or veteran's training, (2) adult vocational agriculture, short courses or regular meetings, (3) mechanical training relatable to agriculture (e.g., welding, carpentry, engine repairing), and (h) other specified training. Of the 1068 farmers responding to this question (additional training), 827 had had additional training and 2&1 had not. Past membership in h-H and.FEA was hypothesized to be positively related to farmers' ability to recognize and give verified examples of -30- the knowledge situations. of the 1028 farmers responding to this question (membership in h-H and FEA), 33 had participated in both the h-H and FFA, 150 had participated in either one or the other, and 8b5 had not been members of either. The farmers with more years of farming experience in operating their own farms were believed more able to recognize and give verified examples of the different knowledge situations. 0f the 1063 farmers responding to the question concerning years farming experience (1) 315 farmers had 10 years or less, (2) 1h5 had 11 to 15 years, (3) 265 had 16 to 25 years, (h) 196 had 26 to 35 YEars, and (S) 1&6 farmers had 36 or more years farm, ing experience. It was hypothesized that younger farmers could recognize and give verified examples of the knowledge situations more often than older farmers. The breakdown of the farmers by age groups is as follows: (1) less than 30 years old -71 farmers; (2) 30 to 3b.? years - 107 farmers; (3) 35 to hh.9 years - 2973 (h) 145 to 511.9 years -- 261; (5) 55 to 61:.9 years -- 23h; and (6) 6S and over -- 9S farmers. As either average gross income, net worth, total assets or total debts increase, it was postulated that the farmers would be more able to recognize and give verified examples of the different knowledge situations. Since average gross income is a 'fairly“ good measure of net worth and total assets, only’average (average of the last 3 years) gross income and total debts were tested for relationships. The ability to recognize and give verified examples of the different knowledge situations was postulated to be positively correlated wdth acres in cropland and rotation pasture. However, the size of farms varied so widely by state and by type of farming, that it was considered impractical -31- to measure the relationships of this variable to knowledge situations. It would have involved the use of index numbers which was not considered ‘worth-while since income should also reflect this variable. In arriving at conclusions, farmers use deduction, induction, and a combination of inductive and deductive thought processes.30 0f the 5&1 farmers asked which method they used, 61 indicated they used mainly de- duction, 127 used mainly induction and 336 indicated that they used both induction and deduction. It was hypothesized that farmers who used mainly the deductive thinking method would be more able to recognize and give verified examples of the abstract knowledge concepts. The same hypothesis was made concerning the natural thinking method of the farmer. It was hypothesized that farmers attending organizational meetings could recognize and give verified examples of the different knowledge situations more often than farmers who did not attend. The organizational meeting referred to includes attending county agent and extension spe- cialist meetings and non-governmental farm organization meetings. .Among the 1075 farmers interviewed, participation was as follows: (1) attend- ance at both kinds of meetings -- 319 farmers; (2) attendance at only county agent and extension specialist meetings -- 151 farmers; (3) attendance at only non-governmental farm organization meetings -- 192 farmers; and h10 farmers did not attend either. .Answers concerning this question were not ascertained from three farmers. The relationships between these independent variables and the farmers' recognition and verification of knowledge situations are tested and presented in Chapter VII. 30The questions concerning the deductive and inductive thought processes were on only 5hl of the 1075 schedules taken. CHAPTER V THE RECOGNITION OF KNOWLEDGE SITUATIONS The general hypothesis concerning the knowledge situations in the IMS can be stated as follows: Farm managers can recognize the five knowledge situations in their own experience. The respondents were asked, for example, "could you please give me some examples of things which you or your family did last year, when you were not completely sure of the out- come, but willing to take the consequences of acting and being wrong?" The farm manager responded in whatever way he saw fit. No choices or examples were given for him.to choose from. This type of question was asked concerning each knowledge situation. The answers or examples given by the respondents were recorded more or less word for word by the interviewer. Afterwards the examples were examined for relevance and consistency. Sometimes situations were con- fused but more often the respondents gave clear, unconfused examples of each knowledge situation as they had encountered in their own experiences. The answers of the respondents were sorted into the following groups: (1) yes answers (farmers who said that they had been in the situation) supported with verified unconfused examples to verify their having en- countered the knowledge situation; (2) those which indicated that the farmer had not encountered the situation; (3) yes answers supported with unverified examples;31 (h) those which confused one knowledge situation 3;A verified example includes what was done and what was involved in doing it. Forexample, a verified example of the risk action situation would indicate what was done and that there was a risk involved in doing it. -33- with another; (5) other answers which could not be classified as examples; (6) those giving no examples but indicated that the situation was not con- fused; and (7) those having no answer. The distribution by answer category is shown in.Table 2. Since it is very difficult to secure data and research information on the different degrees of knowledge held by managers, it is very possible that biases may be introduced. However, in setting up the study much care and effort were taken to minimize these biases by the careful formulation of questions, pretestings and training of interviewers.32 Also, caution was exercised against biases in analyzing the data. The process of securing such data and information on knowledge situations is one requiring a general knowledge of farming, a very high degree of confidence on the part of the interviewee in the interviewer (as is usually the case in getting unbiased data), a close knowledge of the particular problem on the part of the inter- viewer and, most of all, a knowledge of the processes of management on the part of the interviewer. In total there were 1075 farmers interviewed. The breakdown by states in the IMS is given in Table 1, Chapter III. The number and percentage of farmers recognizing each of the knowledge situations are given in Table 2. ' From Table 2, it can be seen that the positive risk action, learning and the certainty situations are the ones for which the most examples were given and which were verified most often while the inaction and positive forced action situations were the ones for which the least examples and verifications were given. It is quite conclusive that all knowledge situations were important but that they varied as to frequency reported. It is interesting to note that the negative risk action situation was among the ones which were encountered most. However, only 35 percent of 325cc page 3U3 for a detailed description of the interviewer training, etc. -3b- .EoHponn nonHoop on» emcee once :nmoH op weHow mm: moans“ one peep new conoapmod was nomeoop page momeHonH novaspHm meHnnmoH may no oHaemwo poHMHnop w .oHdesxo pom .pH weHop :H eo>Hoan mm: pug: one once was was: nononan oHassww poHHHnop ¢.\m m N mm 00H m w Hm mm 00H 533 30.8w opHpHmom Hm a mo ooa m N mN mo ooa saaaapaoo pm a 3 00H 3 a mm mm 00H :oHpomnH oN : oN 00H N m NN so 00H weaanaon mm m mo 00H H 0 mm mm 00H eoHpom meA opHpmmoz hm mH m: OOH m. H 3 mm 8H wages an? opHpHmom :oHpannpmHQ omwveoonom ma NH mmo moo m mm oaN Hod NH; soapoa oooaou opapamom 50H ma aNN oam mm Ha sad mas dNN spewdpaoo mmH mm «mm mmw mH MH :NH 05H Nmm nadwodaH Nd oH HmN mmm NH NN NoN om: oHN meaandoq and 0H oMN own a ms 00: maN mac eoapoa amen opapamoz we mm mm :wH :N N me mu: Ham GOHPow meu mPHpHmom horned. mngdew Ampnsoonm— 0Hmemxfi # oz oz poz can Hapoa scape ooaaeaoo ooaewaoaes \mooamanoe dance Illlumwuopndooqo mo mngmdxv,oz edgeuxo no ucHx and cmnopnwoono mGOHpmava mewmnwwm I'm..- uoHnowopwo moaned mZOHadpaHm mwnmuzbzm BZNMMhhHQ mHH3.mmmezaoozm 924 m0 UZHQZHO>ZH mZOHammDO OH mmmzommmm .mmmZM4m N mnmdy -35- the farmers claiming to have experienced the situation could give verified examples of having actually encountered it. Also, the negative risk action situation was the one confused most by the farmers. It would appear that the definition is not clear but as shall be evident later, the confusion does not necessarily arise from the definition of the situation. Confusion of Knowledge Situations The negative risk action, the positive forced action, and the learning situations were confused.more often than the other knowledge situations. These confusions could arise from a number of sources including inadequate explanations on the part of the interviewers, misunderstanding on the part of the interviewee, and poor or inappropriate definitions of the knowledge situation. However, it is held by the writer that these are among the minor causes for the confusion of knowledge situations. The total confusions and the situations confused are given in Table 3. It is evident that the two situations most frequently confused were usually confused with only two other knowledge situations. The negative risk action was confused most often with the inaction situation, while the positive forced action was confused most often with the positive risk action situation. The possible reasons and explanations for the confusion of knowledge situations will be discussed in Chapter VII. The Importance of Each Knowledge Situation Each of the knowledge situations calls for different action and the use of different principles and strategies to combat risk and uncertainty upon the part of the farm manager. The importance of the knowledge sit- uation is related to the type of action associated with it. In subjective risk situations, the principles of formal and informal insurance can.be employed, of course, the amount of insurance needed de- pends upon the type and seriousness of the risk and the ability of the .mHomvH :pHr pomsmnoo on no: use :oHpmspHn pane moponunH M .uGOHpmpnompo on mopdquqH I .AOHO newsman ON noHszv noHpmspHm onnHm a sane duos omsMeoo has noenmm oco .nmposom .mnoHpmspHm owOonosx unmanned on: newsman OOH mo Hopes a was muons \m mNH m mH NH m on NN NH mN Hmpos mm M I m N I N I HN coHpo4 poonom opHprom 2 N a a I .. N .. a. agofioo “w MH I H I N N N m I eoHpoqu NN I N I a N M m I meHsamoH m: I OH I H MN m N I oneod mem opHpsmoz N H I a I I N I M coHpo< mem opHpHmom Hmpoa :oapod pounce .mozw .mom .Hopew .HoD , wannmoH noHpo¢.mem noHpo¢.me Oomsmcoo opHpHmom hchmpnoO nOHpoqu opprmoz opHpHmom :oHpmspHm owOoHsocx Suzhou 52: ans: 8380.3 omeoazoé MZOHadnaHm mummnzozu mmmao mma mst. ZOHHHDaHm moanzbzu mo<fi UZHmszou \mmmmzm4m mumZDZ m Manda -27- I manager to sustain losses. Informal insurance includes arrangements, generally within the business, to protect against loss. These include (1) discounting returns; (2) liquidity; (3) education; (LI) "excess" horsepower; (5) "excess" feed supply; (6) diversification; (7) main- taining cash reserves and (8) having some unused credit. Formal in- surance includes (1) fire insurance, (2) life insurance, and (3) crop insurance. The learning situation involves the principles of inductive and deductive learning processes. Budgeting, economic principles, continuous function analysis, linear programming, etc. are helpful to farmers in the learning situation. The learning process also requires that the business be flexible or be able to employ the flexibility principle. For the inaction situation, few principles are applicable. Those which do apply deal mainly with determination of the chances which a manager is willing to run in taking action. Here society can employ services which are helpful to the farmers in making decisions. Anything which can be done to reduce the cost of information or increase the value of information will tend to help the manager make a better decision. In the inaction situation, farmers can employ strategy principles. The manager may pick the minimax (the best of the worst outcomes) and try to minimize the maximum loss by choosing the appropriate course of action. However, other people may want to strive for the maximax, where- in, the manager wants to maximize the maximum possible. The manager could select actions such as taking out insurance which would minimize the possible losses that might be incurred. In the subjective certainty situation, the principles of static econo- mics and the principles of budgeting are particularly applicable. In this situation, knowledge is considered so good that the manager can take action -28- without taking the precaution of protecting himself in case he should be wrong. Summary This chapter has been primarily concerned with presenting the gross results from.the survey (INS) on the recognition of the knowledge situa- tions by the farmers and their ability to give verified examples of each situation. The incomplete answer groups were not investigated in this study, however, they seem to substantiate the general hypothesis, but were not consistent and explicit enough to be classified in the verified group. Also, these incomplete answer groups (unverified examples) provide a basis to determine the knowledge situations which are most difficult to understand (ex posts). From this chapter the following conclusions are apparent: (l) in general, the ability of farmers to understand the classification of degrees of knowledge has been confirmed, (2) all the knowledge situations are relevant, (3) the learning, risk action and certainty situations are particularly important, (h) the negative risk action situation is the most difficult for farm managers to understand and easiest to confuse, (5) possibly some other knowledge situations need to be defined. The conclusions reached in this study confirm the results, concerning knowh ledge concepts, of the Kentucky Pilot Study which was conducted by G. L. Johnson in 1951. This analysis has not taken into account every situation a manager encountered. Since a manager could encounter the same situation many times, the order of importance of the situations can not be given. Chapter VI DIFORMRTION AND CONFUSIONS CONCERNING TEE KI‘?O‘.‘:LEDGE SITUA TIOI‘JS The types of information needed by a farm manager in solving his management problems depend, among other things, upon the characteristics of the business which he operates. Thus, the manager may find himself in the position of collecting many types of information simultaneously. In the IMS, types of information are grouped in the following categories: (1) price information, (2) production information, (3) information concerning new developments, (LI) human information, (5) institutional information, and (6) information concerning home technology. In solving a single problem, a manager may find himself drawing on as many as three or four or even all six types of information. Thus, it becomes difficult to call a particular problem a price problem, or a production problem or etc. The DES has furnished some data on the relevance of these different types of information. These are shown in Table 14. In Table h, information concerning production methods was the predominat type of information which farmers indicated they would use in organizing and operating farms for profits. About fifty percent of the time they mentioned production methods (old technology). Another significant feature was that farmers indicated the need for information on new production technolog more often when considering the operation of farms for profit than when considering farm organization. When considering operating the farm for family satisfaction, institutional information was the one pre- dominantly mentioned. -h0- .HoboH £883 98 one. as doggy m .mH 5H3 mONmueweeOmIEO 68.333 one romance 9:83:80 no 303qu annulment...» Omaha. in one. no 303:9: 86ng \m III mom III mom III 4mm Begofifi ugh mo acne—nu 8H mmmfi 8H 03mm OOH Nmmw H309 m.» was m.a . m; a. HH awoaoaeooa ago; Moom dew Hem Nun momN th Haogfivflpaa 93 EN :4 om a.ma mm: ease doc MOH H03” N43 H. J Raw—“9506.8 Gogofidofl to: :4: 03 33 No.9 3% mama sodas 8383; gm mm 0.0N mHm pH 3 cog escaped nonesz endowed $32 and 98m .8952 l o as no figlirlll. wfiflewho nofigomaH Ho and: afieooaeoo an 35‘ JmOH 5" magnum €33.63; no mason—O @3033 be. confine: 0.83 cogaoa no monks .3an Hum 9:. Ho nomm Ho \dmfluononaoo no; no hon—.52 J 393. -hl- OOH m.N m.om 0.0H 2.0 4.:H w.m hHHemm go manenodO OOH m.H H.m :.H H.mH H.:: O.> mpHmoam no ooa a. m.mN a.ma N. o.mm o.H mesaaaamao uanz :OHpooeeoo dH Oopooz nodenHapmHn ommpnooaom OOH H mH ON H Nm 0 :oHpoa.Ooonom opHpHmom 02 H S OH H S S 35380 OOH II mH OH OH no : soaposeH OOH 2 w m m.p mp m.m wanamoH OOH m m 3H H up w :OHpo¢.mem oermmmz ooa N Ha Ha N we a soapo<.amam oaapamoi coHpanwpmHm owmpGOOHom NH: N as NHH a man mm eoapoa.eoonoa oaaoamod aNa Ha mma om m mmm mm spanspaoo NNm H :m mm on omN mH eoHpode map mm Na we as Noe om meaanaoa mam NN Np OHH m me Nm :oHpo<.mem o>Hpemoz Hmm NH OHH . MNH NH map mp eoHpo¢.MmHm o>HpHmom moHdemxo wnHeam .nome HmeOHpsp .moHoeom noHpmepHm newshmm mo oz, meow, IHMmcHI neenm 3oz, .Oohd ooHaa, owpmHemmM QOHpmepHm omeoHaomM coma eH :oHpmenomcH mo many an o>mo mnmewwm moHdemwm Ho Honesz m mamas -lI2— Types of Information by Knowledge Situation The data in Table 5 indicate that production information was men- tioned.more often than other types of information wheanarmers were asked to give examples of each knowledge situation. Mere farmers, however, were able to give examples within the risk action situation, the learning and the certainty knowledge situations than for the other knowledge situations . ‘When the data on each type of information are broken down by know- ledge situations, the picture does not differ from.that secured with the greys tabulations. From Table 6 below, farmers gave more examples in the price category when considering the learning and certainty situations than.when considering other knowledge situations. Mere production information examples were mentioned for positive risk action than for any other knowledge situation. TABLE 6 PERCENTAGES OF FARMERS' EKANPIES GIVEN BY KNOWIEDGE SITUATIONS WITHIN EACH INFORMATION CATEGORY Types of Information Knowledge sti- ‘Home Situation Price Prod. New Develop. Human tution Tech. Positive risk action 21.5 25.0 111.0 22.0 21.0 13.0 Negative risk action lh.0 19.0 6.0 20.0 13.5 29.0 learning 25.0 20.0 h3.o 15.0 13.5 37.0 inaction LI.o 8.0 28.0 7.0 10.0 1.0 Certainty 26.0 18.0 6.0 16.0 28.0 12.0 Positive forced action 9.5 10.0 3.0 20.0 1h.o 8.0 Total {100.0 100.0 100.0 100.0 100.0 100.0 .1, 3.. In the case of new production methods (new developments) more examples were given by farmers illustrating the learning and inaction situations than were given for the other four situations combined. Forty-three per- cent of the farmers giving examples involving new technology were in the learning situation. Human infomation examples occurred more often in giving risk action and positive forced action examples, than in giving examples of the other knowledge situations . Farmers giving examples for the certainty and the positive risk action situation mentioned institutional information more often than when giving other examples. Home technology was mentioned more often for the learning situation than for any other knowledge situation. Confusion of Knowledge Situations ‘91 Type of Information Involved Chapter V discussed the confusion of knowledge situations in a pre- liminary way. Attention is now directed to types of information involved when confusions occurred. The knowledge situations most frequently confused with other knowledge situations were negative risk action and positive forced action. The negative risk action was confused more often with in- action while positive forced action was confused more often with positive risk action. The confusions of knowledge situations by type of information are given in Table 7. The negative risk action accounted for 39 percent of the total confusion. As indicated in Table 7, fifty-three percent of all the confusions involved information on new development (new technology). 0f the two situations confused most, negative risk action and positive forced action, new development information was involved forty-nine and fifty-three percent of the time, respectively. -hh- TABLE 7 THE CONFUSED EXAMPLES OF KNOWLEDGE SITUATIONS BY TYPE OF REFORj'iATION INVOLVED Type of Information Involved Knowledge Situation I “New Insti; ’HOme Not Confused Price Prod. Develop. Human tutional Tech. Ascert. Positive risk action - 3 7 - - - - Negative risk action 7 10 3h 2 ll 5 1 Learning 1 6 16 - l 2 l Inaction 2 l 7 l 3 3 - Certainty 2 l 10 l l 1 - Positive forced action 7 - 21 - 7 h 1 Percentage Distribution Confusion by Total: Examples given 5.2 .7 67.h 1.0 h.1 16.0 1.7 Examples confused 10.5 11.6 53.0 2.2 12.7 8.3 1.7 Theaibove analysis seems to centralize the confusion problem within new technology (new development). The small percentages of confusions involving other types of infbrmation, leads to the conclusion that the main cause of confusion.may be found in the definition of new technology rather than in the definition of the knowledge situations. A further discussion of the confusion of knowledge situations, with particular attention to new technology, is presented ianhapter VII. Summagy Since the schedule only called for respondents to give at least one example of each knowledge situation, we are unable to classify knowledge situations by their relative importance with reference to the ones which farmers encounter most. However, the total number of farmers encountering the different knowledge situations, does give us some idea .115- of which situations are most common among farmers. (See Table 5) .Also, we did not order the decisions which the farmers had to make in terms of their importance. Thus, we cannot say, for example, that the positive risk action and the production information were the most important, but they were the ones indicated.by more farmers. In this chapter we are able to make the following conclusions: (1) all six categories of information used in solving management problems are important,33 (2) production informationuvas mentioned more often than other types of information when farmers were asked to give examples of each knowledge situation, (3) production information was most important in organizing and operating the farm for profits,33 (u) institutional infermation is the predominant type needed when considering operating the farm for family satisfaction,” (5) the risk action, learning and certainty situations were the knowledge situations farmers could give more verified examples of by type of information, (6) and new development was the most frequent type of information involved.in.the confused examples. 33Part of these conclusions were not direct results of this analysis but have only been further expressed. Conclusions 1, 3 and h (above: were given in G.L. Johnson's article, "New Knowledge and Decision Making Processes," presented in the Journal of Farm Economics Vol. hl, (Dec. 1958) - CHAPTER VII ATTRIBUTES OF FARMERS IN DIFFERENT KNOWLEDGE SITUATIONS, SOURCES OF CONFUSION AND REFORMULATIONS In this chapter, the characteristics of farmers able to give verified examples of each particular knowledge situation; are compared. However, the analysis is not subject to generalization in some instances because the variables involved are interrelated and caution must be exercised in estimating the degree of association with any single variable or characteristic. The variables involved often change in the same or Opposite directions, thus, enforcing or off-setting each other. After certain hypotheses and relations are tested, some possible ways of clearing up the tendency of farmers to confuse certain knowledge situa- tions are discussed. While a general ex ante hypothesis was made concerning the ability of farmers to recognize and give verified examples of the knowledge sit- uations, the majority of the hypothesizing consisted of less important ex ante and ex poste hypotheses. Ex pgste hypotheses are those formulated after the data have been collected and inspected in a preliminary way. From.such preliminary inspection, certain hypotheses (2§.22§EE) are set forth and then tested by much more detailed analyses of the data. The obvious disadvantages of using ex poste hypotheses are:. (l) the evidence may not be conclusive and complete since the study was not designed to test such hypotheses and (2) the hypotheses are more or less designed to agree with the data collected instead of the data being collected to test the hypotheses. Offsetting these disadvantages are the following advantages of ex poste hypotheses: (1) they allow for more complete -h7- analysis of the evidence, (2) they help make eXplicit certain indicated relationships, (3) often times, if made and tested, they may save time and money in collecting and analyzing new evidence, (h) they help sub- stantiate other hypotheses made and conclusions reached, and (5) provide explanations for certain relations and indicate possible areas for further research. Statement of Other Ex Ante and Ex Poste Hypotheses The characteristics of farmers which were related to degree of knowb ledge encountered include the following: (1) grade of school completed, (2) farming experience, (3) age of farmer, (h) gross farm income, (5) natural thinking process, (6) organizational associations in which farmers participated, (7) additional training (veteran's training, adult vocation agriculture, etc.), (8) past-membership in h-H and FFA, (9) total debts, and (10) thinking method used. A. The ex ante sub-hypotheses3h concerning the degrees of knowledge are stated as follows: The encountering of the degrees of knowledge and the ability to give verified examples are (l) positively dependent on the grade of school completed (2) positively related to years of farming experience (3) negatively related to the age of the farmer (h) positively related to gross farm income (5) related to the natural thought process, i.e. whether it is most natural for the farmer to reason inductively, deductively, or a combination of both (6) positively related to the number of organizational affiliations 3hFor a more detailed discussion of these hypotheses see pages 29 to 31 of this thesis. -h8- (7) positively related to the additional training received (8) positively related to membership in the h-H and FEA (9) positively related to total debts (lO) positively related to the thinking method used (in the fol- lowing order; induction, both, and deduction). The characteristics of farmers encountering one degree of knowledge are not necessarily the same as those encountering another degree of knowledge. The chi-square test for independence was employed to test these variables for independence with regard to the ability of farmers to recognize and give verified examples of actually encountering the knowe ledge situations. The test of the hypotheses under (A) above by each knowledge situation yielded the following list of variables as having dependent (at the 10 percent level) relationships with ability to recog- nize and give verified examples of actualLy encountering the indicated knowledge situations (where f stands for the phrase "a function of" and - stands for a relationship with sign other than stated in the hypotheses above): I. pgsitive risk action - f (education, years farming experience, natural thinking process, association with organizations, total debts, and thinking method used). 2. negative risk action - f (education, age, association with organizations, additional training, total debts, and thinking method used). 3. learning situation - f (education, - years farming experience, age, gross income, association with organizations, additional training, membership in FFA and h-H, and thinking method used). h. inaction a f (education and thinking method used). 5. .gertainty'- f (education, years farming eXperience, gross income, _.9_. additional training, membership in FFA and h-H, and total debts). 6. pgsitive forced action = f (education, years farming experience, gross income, association with organizations, additional training, total debts and thinking method used). B. The characteristics of those farmers who did not encounter the knowledge situations were essentially the inverse of those who did encounter and give verified examples, i.e. where the rela- tionships above are positive, these would be negative, etc. C. Ex posts, it was hypothesized that the type of information_given under each knowledge situation is independent of education, years farming eXperience, age, gross income, etc. This hypothesis is derived from a more fundamental proposition that the type of in- formation needed is determined by the problem, not the character- istics of the farmers. Thirtyhsix chi-square tests for indepen- dence were computed in comparing the characteristics of farmers with the type of information given, only one of which yielded a significant difference or dependence (at the 5 percent level). This variable was gross farm income and the knowledge situation involved was negative risk action. However, it is still con- cluded that the type of information given is independent of the characteristics of farmers under consideration. This is possible because over twenty percent of the components or expected values in the one table (which yielded statistical significance) tested were less than five. Thus, we can conclude that this particular test is unreliable.35 35If any of the expected values in the computation is one or over 25 percent of the eXpected values are less than 5, the chi-square test is unreliable. For further explanation, see W. J. Dixon and K. J. Massey, Jr., Introduction to Statistical Analysis, second edition, MbGrawaHill 00., New'fork, 19§73 p. 222. Also, see pages 106.7 in Statistical Inference by H. M.‘Walker and J. Lev, Henry Holt and Co., New-Yorkj”l953. ~50- D. Ex poste, it was hypothesized that the inability of a farmer to recognize one knowledge situation was independent of his in- ability to recognize another. Also, it was hypothesized that the ability of a farmer to_give a verified example of one know- ledge situation was independent of the ability to give a verified example of another situation. The tables for six of the ten inde- pendent variables were tested. The summarized results of these tests are given below in Table 8 both for farmers who did not indicate that they had encountered the knowledge situations and for farmers giving verified examples. Twelve chi-square tests were computed for independence, of which, there was not a case of dependence, (at the 5 percent level). Thus, the above hypotheses are accepted as confirmed. Characteristics of the Farmers Confusing Knowledge Situations The characteristics of those farmers confusing the negative risk action and the positive forced action, the two most commonly confused situations, have been investigated with respect to the ten independent variables. The results of such investigations are summarized in Table 9. It can.be concluded that the characteristics of farmers who confused a knowledge situation are not practically different from those who gave unconfused verified examples though they are significantly different from a statistical standpoint in the case of age and past membership in h-H and ERA for the positive forced action knowledge situation. It must be remembered that it is difficult to attribute cause of a particular incident to any one variable. The age of those farmers confusing the positive forced action knowledge situation were statistically signif- icantly different from those farmers giving unconfused verified examples. The age of a farmer is closely correlated with education and years farming experience; thus the variable may be, in effect, a composite variable 09mm was and ma essences Hgofiesflnwwha H93 hm o.c 2 scenes antenna 80% moNH dowfl m4” mEOOGH moém. 0.2” and mm mm mwém «.3 m.om mm communes manna 34m «.3 méa om dogma: .3va panama m an newsflash _ Sac» on?» a , o 0 com gasob cocoaamo aooooflm 33.3.“an ”swam assesses easements scz can acts as mfimfiuo one wfinfidwooom mnofimm page $.38ch whoefiwm moonwon peopcoaoonH mZOHBSEm go mo “SE“. 83. ZOHBSfiHm m§.7..ozv~ flzo mo gaming Emu—”ENE E0 92¢ HNHz—uoumm on. HBHHHmSAH 84¢ HHHHHmd .mmagh ho mmSmHEnq BE. mo zomHfifiaoo d mama. -52 .. pomofiwflcmflm hHHmoflpmapmpm ohms moanmwnme omega * monsoHMfieme mo Ho>mH peoohom m one pa\m ahm and no.m a some scenes age and and 8.: in fines asses amnm amoe and N am can m..: an.” assented: sea 2. cm. H wfifieh 38333 .8; and an m anaesthetised .ssnso elm cm: oo.m m ecanca nfiafifi assess .5; +3. 0H.H m. anocH 83H e34.” RA m ems 8.3 ads 93 m confiscate madame an; 34 and 4 segments odaw> osam> NN oon> soooohm adamant» \lcohaovmm oo930300 oopcaemo we no noapo<.ooonoh o>HpHmom eowpo<.xmflm o>apmmoz moonwom ofipmflnopoeameo manna”? QmHmHE QMmDB,HooH/.._ 056 033 whoma 29H; ZOHBdBHm munufliozm UZHmbmzoo mmmzmam MM-B ho mOHBmHmEBo¢M¢mo HEB mo ZomHmdmzoo d m Em