The object of this study was to determine whether statistical decision theory, or a special application of it, the theory of signal detection, could be of value in accounting for the behaviour of subjects in a perceptual decision-making task. The amount of information in these tasks was varied to see if the theory could predict changes in subject performance. Five subjects were required to distinguish between fifty percent time compressed recordings of the stimulus words "commination" and "comminution” embedded in "white" noise. Under one treatment, compression was gained by discarding many small letter segments while in the other this same compression value was obtained by discarding a few large letter segments. It was hypothesized that large-discard- interval compression would be more detrimental to stimulus intelligibility than small-discard-interval compression. Five other subjects were asked to distinguish between the two noise-embedded stimulus words which had been time-compressed sixty and seventy-four percent. It was predicted that sixty percent compression would be less detrimental to the intelligibility of the stimulus words than seventy-four percent compression. Concurrently, in both groups, an attempt was made to manipulate the degree of cautiousness or decision criteria of all ten subjects. Such manipulation was attempted in order to permit the separation of each subjects' actual sensitivity from each's variable decision criterion. This manipulation involved varying the costs and fines associated with correct and incorrect decisions as well as the probabilities of each stimulus word's occurrence. Large-discard-interval compression was found to be less detrimental to intelligibility, as inferred from subject performance, than small-discard-interval compression. This finding was contrary to the first hypothesis. Sixty percent compression, as predicted, was less detrimental to intelligibility than seventy-four percent compression. It was observed that the theory of signal detection permitted separation of each subjects' sensitivity from his monetary degree of cautiousness. This cautiousness was also found to be accessible to manipulation. It is suggested that since the approach of statistical decision theory detected changes in subject performance in response to varying amounts of information, it can be profitably applied to the study of perception. / Arts, Faculty of / Psychology, Department of / Graduate

The problem addressed here is the examination of the dilemmas of decision making in different substantive contexts. The contexts include peasant farmers deciding whether or not to accept an agricultural innovation, sophomore students gambling, and fishermen on British Columbia's West coast making up their minds where they intend to fish. The major conclusion is that the structure of decision making is a constant cross-cultural variable. This implies that socio-cultural factors can most profitably be viewed as a framework within which a similar structure of decision making occurs. From this consideration it follows that decision making is a basic building block for the study of social behavior. Identification of the basic structure of decision making is in terms of a theory of risk taking which relates the type of decision strategy used in any situation to considerations of the resources, information and utilities that particular individuals possess with regard to the event the decision is about. From an initial substantive concern with Third World farmers deciding to adopt or reject agricultural innovations I generalise to a number of statements about individuals and risk. Risk taking refers to behavior in situations where there is a desirable goal and a lack of certainty that it can be achieved with attendant possibilities of loss. Three main sources are used to test the assertions about risk taking — first a laboratory experiment, then fieldwork and finally secondary sources. The argument made to justify this procedure is that these situations constitute particular empirical settings in which the propositions about risk taking could legitimately be tested. The argument rests on the assumption that the same scope conditions are met in each substantive setting. The scope conditions considered here place an individual decision maker within parameters of resources and subjective utility with regard to some outcome, information and incentive conditions for any risk. The propositions predict the type of decision strategies that would be employed for given values of the above parameters. This level of abstraction, which is not tied to situational boundaries is, I submit, a necessary prerequisite for effective cross-cultural analysis. Thus my thesis is not about peasant farmers, or fishermen or gambling, the work attempted here is concerned with individuals and risk. The tools used draw upon a tradition of model building extant in economics with reference to decision theory. Thus the work attempted here is part of the growing formal tradition in economic anthropology. The model built was not a perfect fit with data but adequate enough to give one confidence in the set of methodological assumptions which were fundamental to its construction. Also the testing procedure employed has implications for the manner in which anthropologists may conduct enquiry, as it establishes that laboratory contexts are as legitimate a source of verification as field contexts. It is from these two considerations that I offer a test case for methodology in economic anthropology. / Arts, Faculty of / Anthropology, Department of / Graduate

This thesis developed a pragmatic approach to the optimization of business decisions in the chloralkali industry utilizing basic marginal analysis concepts. It attempted to meld accounting principles and techniques with economic concepts for the purpose of establishing a rationale and a workable model for guiding and evaluating everyday business decisions. The approach concentrated on short run decision making and did not get involved with capital and other variables which were fixed in the short run. Many attempts to apply marginal analysis to the real world have had only limited success. The concepts as outlined in this thesis used a particular empirical example to circumvent some of the inherent difficulties associated with marginal analysis. The emphasis in this study was to develop a useful business tool and not to fit theoretical economic concepts into the real world. The special requirements for this model were a capital intensive multiproduct company with a vertically integrated production line. Each product manufactured should have two outlets; a) to be used as raw materials to upgrade the product by further processing, or b) to be sold in the market place. Consequently, the product lines will be developed from relatively few raw materials. This is the situation in most petrochemical and inorganic chemical plants, oil refineries, and could even be applied to entire industries such as forest products, and pulp and paper. Although the chloralkali model's construction were discussed, the model itself was not included in the thesis. The particular model used was large and complex, specially adapted for the firm, and as a result contained confidential information and relationships which were difficult to disguise. The empirical example assisted in the development of the essential concepts but did little to contribute to the proving of the hypothesis. / Business, Sauder School of / Graduate

A new approach for making management decisions about the logging of forested streambanks is presented in this thesis. For several years administrative guidelines or regulatory clauses have required that narrow strips of vegetation be left along all streambanks after logging but experience with these guidelines or regulatory clauses has not always been satisfactory. In many cases the type of strip that is left is not well suited to the particular streambank site or to the needs of the users of that site and a new procedure for making streambank management decisions is clearly needed. The decision-making procedure developed and tested in this thesis provides a routine, consistent method for site-specific decision-making on any Vancouver Island streambank site slated for logging. It includes a range of possible management alternatives and a method for determining which alternative is appropriate to a given site. A checklist is used to assess a number of physical characteristics of the stream and streambank and to determine the resource uses of the site. The value of the vegetation along the streambank to the users is then identified and from this environmental basis, the width and type of vegetated leave strip best suited to the integrated use management of the site is indicated. The procedure was field tested on seven different streambank sites on Vancouver Island. Checklists were completed by individual participants and decisions about the width and type of strip to be left were made according to the procedure. Written qualitative comments were solicited. Analysis of the field test results indicated that there were a number of shortcomings in the checklist and in the procedure generally and revisions and improvements are suggested. There was however sufficient consistency in the results to indicate that the procedure could be used routinely and consistently by field level personnel and would improve streambank management decisions. The qualitative reaction to the procedure was generally favourable. It was felt to be a useful and valuable innovation and several participants indicated that they would be prepared to use it on an operational basis. It is concluded that a useful procedure for making site-specific integrated use decisions has been developed for forested streambanks. Recommendations are made for further improvement of this procedure and for its application to other resource management problems. / Arts, Faculty of / Geography, Department of / Graduate

This dissertation consists of two parts. Part I contains the statements and proofs of two representation theorems. The first theorem, proved in Chapter 1, generalizes the quasilinear mean of Hardy, Littlewood and Poly by weakening their axiom of quasilinearity. Given two distributions with the same means, quasilinearity requires that mixtures of these distributions with another distribution in the same proportions share the same mean, regardless of the distribution that they are mixed with. We weaken the quasilinearity axiom by allowing the proportions that give rise to the same means to be different, This leads to a more general mean, denoted by M[sub=αФ], which has the form: M[sub=αФ] = Ф⁻¹(ʃ[sub=R] αФF/ʃαdF), where α is continuous and strictly monotone, a is continuous and strictly positive (negative) and F is a probability distribution. The quasilinear mean, denoted by M[sub=Ф], results when the a function is constant. We showed, in addition, that the M[sub=αФ] mean has the intermediate value property, and can be consistent with the stochastic dominance (including higher degree ones) partial order. We also generalized a well known inequality among quasilinear means, via the observation that the M[sub=αФ] mean of a distribution F can be written as the quasilinear mean of a distribution F[sup=α], where F[sup=α] is derived from F via a as the Radon-Nikodym derivative of F[sup=α] with respect to F. We noted that the M[sub=αФ] mean induces an ordering among probability distributions via the maximand, ʃ[sub=R] αФF/ʃαdF, that contains the (expected utility) maximand, ʃ[sub=R] αФF, of the quasilinear mean as a special case. Chapter 2 provides an alternative characterization of the above representation for simple probability measures on a more general outcome set where mean values may not be defined. In this case, axioms are stated directly in terms of properties of the underlying ordering. We retained several standard properties of expected utility, namely weak order, solvability and monotonicity but relaxed the substitutability axiom of Pratt, Raiffa and Schlaifer, which is essentially a restatement of quasi-linearity in the context of an ordering. Part II of the dissertation concerns one specific area of application decision theory. Interpreting the M[sub=αФ](F) mean of Chapter 1 as the certainty equivalent of a monetary lottery F, the corresponding induced binary relation has the natural interpretation as 'strict preference' between lotteries. For non-monetary (finite) lotteries, we apply the representation theorem of Chapter 2. The hypothesis, that a choice agent's preference among lotteries can be represented by a pair of α and Ф functions through the induced ordering, is referred to as alpha utility theory. This is logically equivalent to saying that the choice agent obeys either the mean value (certainty equivalent) axioms or the axioms on his strict preference binary relation. Alpha utility theory is a generalization of expected utility theory in the sense that the expected utility representation is a special case of the alpha utility representation. The motivation for generalizing expected utility comes from difficulties it faced in the description of certain choice phenomena, especially the Allais paradox. These are summarized in Chapter 3. Chapter 4 contains the formal statements of assumptions and the derivations of normative and descriptive implications of alpha utility theory. We stated conditions, taken from Chapter 1, for consistency with stochastic dominance and global risk aversion and derived a generalized Arrow-Pratt index of local risk aversion. We also demonstrated how alpha utility theory can be consistent with those choice phenomena that contradict the implications of expected utility, without violating either stochastic dominance or local risk aversion. The chapter ended with a comparison of alpha utility with two other theories that have attracted attention; namely, Allais' theory and prospect theory. Several other applications of the representation theorems of Part I are considered in the Conclusion of this dissertation. These include the use of the M[sub=αФ] mean as a model of the equally-distributed-equivalent level of income (Atkinson, 1970), and as a measure of asymmetry of a distribution (Canning, 1934). The alpha utility representation can also be used to rank social situations in the sense of Harsanyi (1977). We ended by pointing out an open question regarding conditions for comparative risk aversion and stated an extension of Samuelson's (1967) conjecture that Arrow's impossibility theorem would hold if individuals and society express their preferences by von Neumann-Morgenstern utility functions. / Graduate and Postdoctoral Studies / Graduate

The information seeking behaviour of decision makers is an important aspect of the rational decision making process. This study examined the information seeking behaviour of students of educational administration in an attempt to more fully understand the decision making process and thus provide a basis for improving rational decision making. A statistical decision game was devised which utilized a Bayesian norm. The value of information to a decision maker was calculated from the game and then compared to the value defined by the Bayesian norm. The sample consisted of twenty-nine part-time students enrolled in winter session courses in the University of British Columbia (U.B.C.) Department of Educational Administration. This sample included students with various amounts of completed class work and educational administrative experience. Three independent variables were defined. The first variable, administrative experience had two levels. One level consisted of those with at least one year of administrative experience, the other consisted of those with no administrative experience. The second variable, graduate level training, had two levels. One level consisted of those who had completed at least three units of graduate level course work in educational administration. The other consisted of those who had not completed any coursework. The third variable, consisted of three divisions with two levels each and introduced varia tion across the eight games. The dependent variable, information seeking bias, was defined as the difference between the quantified value of information to the decision maker and that value ascribed by the Bayesian norm. The sample was tested in separate groups ranging in size from three to six. Each person in a group viewed the same sequence of eight games, corresponding to the eight treatment levels of factor C; however, each group had a different randomly determined sequence of games. The following five research questions were examined: (1) Do students of educational administration tend to seek the same amount of information as purely Bayesian players in a programmable decision situation? (2) Do students of educational administration with administrative experience behave differently than students with no administrative experience in their tendency to seek information? (3) Do students of educational administration who have completed three or more units of graduate level training in educational administration behave differently than students with no training in their tendency to seek information? (4) Among students of educational administration does varying the prior probabilities and payoff parameters in the statistical decision game effect the tendency to seek information? (5) Does information seeking bias change over the sequence of games? Does learning take place? The experimental design was based on Winer's (1971) three factorial repeated measure design (case II). In research question four a Scheffe Test was to be employed if the previous analysis indicated certain combinations of treatment levels in factor C had resulted in significant findings. Research question five was analyzed by graphing the information seeking bias for the various subject groups over the sequence of games. It was found that the group as a whole showed a strong tendency to seek more information than the Bayesian optimal. The hypothesis that the mean information seeking bias between the two levels of experience and training was the same could not be rejected. However, the interaction effects between factors A (experience) and C (repeated measures) and between factors B (training) and C were found to be significant at the .10 level. Finally, strong learning curves were noted for all subject groups. As the sequence of games progressed, the mean information seeking bias for all subject groups more closely approximated the Bayesian optimal strategy. It was noted, however, that experienced administrators tended to approach the Bayesian optimal strategy at a faster rate than the non-experienced group. The study concluded by recommending that programme developers in educational administration should attend to (1) the previous training and administrative experience of students and (2) the conditions under which decisions are made, when planning for future programs on decision making. / Education, Faculty of / Educational Studies (EDST), Department of / Graduate

This research investigated a number of hypotheses concerning the form of social influence between factions in small group decision making. In an effort to avoid the bias which may result from the use of confederates in such research, the six person groups used in this study were composed entirely of experimentally naive subjects. Five specific hypotheses were investigated: a) that the influence of a faction depends upon the extent to which they are perceived as cohesive and consistent by the others (i.e., on their apparent solidarity); b) that as the size of a faction increases, that faction will be perceived by the others as being both more competent and less confident with respect to the issue at hand; c) that the faction which wins the first convert will exert more influence on the final decision; d) that disagreement will be solved by compromise; e) that the influence of a faction is a function of the frequency of communication by that faction. The problem discussed was a labour-management dispute. Eight to ten volunteer subjects whose responses to a wage settlement item on an opinion survey fell into either the range of 87o-127o or 307o-407o were scheduled to participate in a group, with this scheduling manipulated in such a way that there was a majority of "high" subjects assigned to each group. On arriving for the experiment, subjects were given a brief description of a labour dispute and asked to indicate their personal belief regarding the proper percentage wage increase. Six of the subjects were then selected to serve as the discussants, and the remaining subjects were asked to act as observers. Because the survey opinions proved to be unreliable indicators of the responses to the issue used in the experiment, the initial groups of eight to ten subjects often did not have a distinct "high" -"low" split, and for this reason the selection of the six discussants was done so as to maximize the polarization within the discussion group. The discussants were then given twenty-five minutes in which to simulate an arbitration of the dispute and to reach an agreement. During the discussion, the observer subjects and the two experimenters observed the group from an adjacent room through an observational glass. The discussion was recorded on audio tape and the two experimenters also independently coded each comment made by the subjects in terms of which person was speaking, to whom she spoke, and whether her comment favoured a "high" or "low" decision. When the group reached consensus (or at the end of the allotted time), all subjects completed a questionnaire concerning the course of the discussion, characteristics of the discussants, and opinions on a similar case. Fifteen discussion groups were used in the study. The results indicate: a) that compromise was not the means by which decisions were reached; b) that while the final decision did not always favour the faction which won the first convert, that faction was generally more influential; and c) that the other three hypotheses were unsupported. There are certain limitations to the generalizability of these data, based upon the combination of the small sample size and the amount of between group variability. The implications of these factors are discussed in detail, both with respect to the reliability of the obtained results and the problems for future research. / Arts, Faculty of / Psychology, Department of / Graduate

No description available.

Managers are increasingly faced with making complex decisions in turbulent organisational environments. This has led to greater information processing demands. Increasingly organisations try to deal with this in such a way that many of these decisions are now made in a group environment. The increase in group decision making has generated a corresponding intensification in the interest in options available to support such decision making. One such approach is a Group Decision Support System (GDSS) referred to as Decision Conferencing. However, Decision Conferencing rests on the unsupported key premise that the computer modeling, which forms an intrinsic part of the process, leads to shared understanding and commitment - the stated goals of the process. The application of Decision Conferencing to important organizational issues continues, yet prior to this study its fundamental premise was both empirically unsupported and potentially under-theorised. This theory-building research demonstrates that the interface between these concepts is more complex than the literature suggests and that the concepts themselves are problematic. Shared understanding is essentially a dependent variable, with factors such as comprehension of the modeling process impacting on the degree to which this is developed. In addition, many aspects of commitment fall outside of the domain of the Decision Conference workshop e.g. the individual’s sense of responsibility and degree of commitment to their profession. The idea of commitment appears to fall more into the arena of managerial responsibility and change management and it is partly how the outcomes are managed after the Decision Conference which will be crucial to their implementation. / Within this study it appears that the most a Decision Conference can offer is the ‘buy-in’ or constructive involvement of the individual participant; the assurance of an unassailable case to which all participants have contributed, for the adoption of the outcomes; and the confidence in the outcomes that this brings. All of this suggests that a higher order goal which subsumes these factors should be considered when re-conceptualising the Decision Conferencing experience. It is suggested here that Decision Quality is a more appropriate goal for the Decision Conferencing process. In essence this is an expansion of the existing ‘best bet’ concept already endorsed in the Decision Conferencing literature. The thesis presents a number of conditions for assuring decision quality e.g. a democratic environment for decision making; mutual respect and an encouragement of diversity. It is also argued that it falls to the facilitator to encompass all of these factors. Given the above, it is also suggested that it is appropriate to consider an alternative conceptualization of Decision Conferencing which facilitators of public sector groups might adopt. This revised conceptualization is drawn from complexity theory. Incorporating the findings from this study a more strongly theorised facilitation approach, entitled Quality Facilitation Practice (QFP) has been developed. Taking into account all of the above a revised model for Decision Conferencing in the public sector is presented, incorporating both QFP and the higher order goal of Decision Quality.

No description available.