Thesis (Ph. D.)--University of Wisconsin--Madison, 1966. / Typescript. Vita. Description based on print version record. Includes bibliographical references.

Thesis (M.S.) in Resource Ecomonics and Policy--University of Maine, 2002. / Includes vita. Includes bibliographical references (leaves 76-79).

Thesis (Ph. D.)--University of Wisconsin--Madison, 1974. / Typescript. Vita. eContent provider-neutral record in process. Description based on print version record. Includes bibliographical references.

No description available.

The purpose of this study is to determine which of three forecasting methods provides the most accurate short-term forecasts, in terms of absolute and mean absolute percentage error, for a unique set of data. The study applies three forecasting techniques--the Box-Jenkins or ARIMA method, cycle regression analysis, and multiple regression analysis--to quarterly sales tax revenue data. The final results show that, with varying success, each model identifies the direction of change in the future, but does not closely identify the period to period fluctuations. Indeed, each model overestimated revenues for every period forecasted. Cycle regression analysis, with a mean absolute percentage error of 7.21, is the most accurate model. Multiple regression analysis has the smallest absolute percentage error of 3.13.

This dissertation is two essays in business decision-making. The first essay is motivated by recent field evidence suggesting significant reliance on conventional techniques (e.g. NPV and DCF) without assessment of the decision profile - its degree of uncertainty, ambiguity and knowledge distribution. However, without knowing the decision profile, the chosen decision might not be appropriate given the decision situation. Therefore, essay 1 develops a multi-faceted conceptualization of the decision profile and provides a prescriptive model for choosing appraisal methods based on this profile. Specifically, it prescribes the limited use of conventional methods to low ambiguity and uncertainty situations and using decision trees, real options, scenario planning and case-based methods as the level of uncertainty increases. In high ambiguity situations, however, the only viable approaches are case-based methods which do not have perfect information assumption that conventional alternative methods do. Case-based methods have been supported theoretically in case-based decisions and case-based reasoning literature but lags in its use in business decision-making. Possible reasons for this include a lack of concrete applications and developments of major concepts such as its case memory, similarity and prediction functions. Therefore, essay 2 proposes a model of case-based decisions called similarity-based forecasting (SBF) and applies it to a high uncertainty and ambiguity situation -- namely forecasting movie success. In doing so, it outlines operational definitions of the memory, similarity and prediction functions and, based on data from the entertainment industry, provides empirical support for the hypothesis that case-based methods can be more accurate than regression forecasting; both SBF and combined SBF-regression models were able to predict movie gross revenues with 40% and 50% greater accuracy than regression respectively. This essay concludes with a discussion of some possible directions for future research including applications using data from other domains and settings, testing the boundary conditions for which the SBF approach should be applied, experiments using SBF under uncertainty and complexity manipulations, and 'time stamped' comparisons with predictions made using information markets (e.g. Hollywood Stock Exchange).

No description available.

This dissertation is two essays in business decision-making. The first essay is motivated by recent field evidence suggesting significant reliance on conventional techniques (e.g. NPV and DCF) without assessment of the decision profile - its degree of uncertainty, ambiguity and knowledge distribution. However, without knowing the decision profile, the chosen decision might not be appropriate given the decision situation. Therefore, essay 1 develops a multi-faceted conceptualization of the decision profile and provides a prescriptive model for choosing appraisal methods based on this profile. Specifically, it prescribes the limited use of conventional methods to low ambiguity and uncertainty situations and using decision trees, real options, scenario planning and case-based methods as the level of uncertainty increases. In high ambiguity situations, however, the only viable approaches are case-based methods which do not have perfect information assumption that conventional alternative methods do. Case-based methods have been supported theoretically in case-based decisions and case-based reasoning literature but lags in its use in business decision-making. Possible reasons for this include a lack of concrete applications and developments of major concepts such as its case memory, similarity and prediction functions. Therefore, essay 2 proposes a model of case-based decisions called similarity-based forecasting (SBF) and applies it to a high uncertainty and ambiguity situation -- namely forecasting movie success. In doing so, it outlines operational definitions of the memory, similarity and prediction functions and, based on data from the entertainment industry, provides empirical support for the hypothesis that case-based methods can be more accurate than regression forecasting; both SBF and combined SBF-regression models were able to predict movie gross revenues with 40% and 50% greater accuracy than regression respectively. This essay concludes with a discussion of some possible directions for future research including applications using data from other domains and settings, testing the boundary conditions for which the SBF approach should be applied, experiments using SBF under uncertainty and complexity manipulations, and 'time stamped' comparisons with predictions made using information markets (e.g. Hollywood Stock Exchange).

Thesis (Ph.D.) - Carleton University, 2002. / Includes bibliographical references (p. 253-260). Also available in electronic format on the Internet.

Thesis (Ph. D.)--University of California, San Diego, 2000. / Vita. Includes bibliographical references (leaves 65-68).