USING AN UNCERTAINTY MODEL TO ASSESS SALES RESPONSE TO ADVERTISING*

Implications of a probabilistic “random walk” model of incremental sales response to advertising are developed for various timing patterns of advertising expenditures. Maximum likelihood procedures for assessing advertising effectiveness and for estimating a decay (forgetting) rate are developed and applied to artificial data of known configuration and are used to assess the impact over time of a brochure program on mail-order sales. Results are also compared to those from alternative models involving various lag patterns in advertising effects.     

AN APPROACH FOR IMPROVED INTERPRETATION OF MULTIVARIATE ANALYSIS *

This paper proposes a method, canonical rotation analysis, which facilitates the substantive interpretation of results in multivariate analysis. Canonical rotation analysis is developed as a model which integrates multivariate least squares approaches and the varimax rotation criterion. The generalized applicability of the model to canonical correlation, multiple discriminant analysis, and multivariate analysis of variance is developed. The advantages and limitations of canonical rotation analysis are discussed and illustrated in the context of an industrial marketing research problem.     

AN IMPLICIT/EXPLICIT APPROACH TO MULTIOBJECTIVE OPTIMIZATION WITH AN APPLICATION TO FOREST MANAGEMENT PLANNING*

Implicit utility/value maximization and explicit utility/value maximization are identified as two major classes of multiobjective optimization methods. Explicit methods have the advantage of being able to fully exploit the power of existing mathematical programming algorithms. A disadvantage is the high information burden they place on the decision maker. Implicit (i.e., interactive) methods have complementary strengths and weaknesses: they require less extensive information but do not lend themselves as easily to use with optimizing algorithms. We develop a hybrid implicit/explicit approach that attempts to combine the advantages of both by embedding within the implicit method a procedure that periodically formulates an approximate explicit representation of the multiobjective problem and then solves it optimally without user interaction. Operationally, this requires the frequent solution of two nonlinear programs. We also report on the implementation of this method in a forest management decision support system. This is a completely microcomputer-based implementation currently undergoing field testing for use in planning the timing and intensity of timber harvests on non-industrial forests in the southeastern United States. The system has been selected as a replacement for an earlier multiobjective program (Harrison and Rosenthal [28]) used by over 1,800 landowners.     

AN OPERATIONAL APPROACH TO THE EVALUATION OF LEASES BY HOSPITALS*

The evaluation of the lease versus purchase alternative by not-for-profit hospitals presents a problem that stems from the interaction between investment and financial decisions. This paper provides an analytical framework that resolves this issue by neutralizing the impact of the lease contract on the hospital's financial structure and debt capacity. The formulation incorporates the special operating characteristics of not-for-profit hospitals: namely, they are not subject to federal income tax, and part of their revenue is generated through cost-based reimbursement.     

AN ADAPTIVE APPROACH TO TIME-SERIES FORECASTING

This paper extends the applicability of the Carbone-Longini adaptive estimation procedure (AEP) to time-series forecasting. Comparisons with adaptive filtering, the Box-Jenkins methodology, and multiple regression analysis as it applies to time-series analysis are provided. Specific time-series data examined by Box and Jenkins and Box and Tiao constitute the basis for these comparisons. The analysis of the results indicate the robustness and performance superiority of the simple distributive-lag forecast model coupled with the concept of adaptively “tracking” rather than “fitting” historical data.     

INTERVENTIONS INTO GROUP PROCESS: AN APPROACH TO IMPROVING THE EFFECTIVENESS OF GROUPS*

Some new directions for research on group effectiveness are proposed. These include: (a) explicit recognition of group interaction process as the moderator of input-output relationships in groups; (b) research which focuses on the functions of group interaction in influencing group performance; and (c) use of experimental interventions which create new, non-typical patterns of interpersonal behavior in groups as an approach to studying group effectiveness. The results of two studies based on these proposals are summarized. One study examines the effectiveness of implicit vs. explicit discussion of group task “strategies”; the other addresses the effects of implicit (“traditional”) vs. explicit (“adaptive”) approaches to the maintenance of internal social relationships in work groups. Implications of the research are drawn both for future research on group effectiveness, and for the design of interventions aimed at improving the performance of on-going groups in organizations.     

A MULTIATTRIBUTE DECISION-MAKING APPROACH TO THE SELECTION OF AN AUXILIARY DEVICE FOR ICEBREAKERS*

This paper describes the use of multiattribute decision making by the U.S. Coast Guard when choosing the most appropriate auxiliary device to use on an icebreaker. Five different missions of icebreakers are defined, and the objectives and attributes that describe the effectiveness of each auxiliary device in accomplishing these missions are established. For each geographical area of interest, the missions are weighted using the analytical hierarchy process. Also, group utilities are developed for attribute vectors in order to incorporate the judgments of different icebreaker operators.     

AN INTEGRATIVE APPROACH TO PROCESS PARAMETER SELECTION

This case study of a process improvement project from the Hewlett-Packard Analytic Products Group illustrates a number of important approaches for operations strategy. First, we show how process control in capillary tube manufacturing led to a significant improvement in tubing strength, a quality measure that relates to the frequency of tube defects at a particular stress level. Next, we show how one particular approach used for experimental design and process improvement was considerably more economical than another standard method. In conclusion, we link the process characteristics such as defect frequency to strategic measures such as company revenue and investment to underscore the strategic importance of process control.     

AN ASSIGNMENT APPROACH TO CHOOSING R&D EXPERIMENTS

The criteria upon which R&D experiments are evaluated are subjective; therefore in practice it is extremely difficult to obtain a series of weights so that the trade-offs between the various criterion can be analyzed. The problem is compounded when the experiments represent basic research and no prior history is available for utility estimation. This paper presents a graph-theoretic interpretation of the selection process based only on the rank-order of the alternatives thus eliminating the need for utility estimation. Index numbers are then obtained as the solution of an integer program. An application and the theoretical foundations of the procedure are discussed.     

CONSISTENT RANKINGS BASED ON TOTAL AND DIFFERENTIAL AMOUNTS UNDER UNCERTAINTY*

Rankings of decision alternatives based on total and on differential monetary amounts may in general be inconsistent under uncertainty. In the following cases, either approach is valid and hence yields consistent rankings: (i) with expected values, provided that the differential amounts have been coherently determined and are constant within states (but may differ across states); (ii) with exponential utility functions if the shared and differential amounts are statistically independent; or (iii) in a market valuation context, assuming diversification and implicit separate market values for differential and shared amounts.     

A BAYESIAN APPROACH TO PLANT-LOCATION DECISIONS*

The paper addresses the problem of plant location in the formal context of decision making under uncertainty and presents a framework employing Bayesian analysis in the collection and assessment of information. As a general model, the Bayesian approach is shown to subsume two practical approaches common to plant-location literature: satisficing and spatial hierarchy of plant-site characteristics.     

AN EXPERIMENTAL EVALUATION OF THE RELATIONSHIP OF UNCERTAINTY IN THE ENVIRONMENT TO INFORMATION USED BY DECISION MAKERS*

This paper discusses the relationship between uncertainty in a decision making environment and the desired characteristics of information used for decision making. The work is aimed at providing a better understanding of the variables that affect the design of Management Information Systems. An experiment was conducted using an inventory simulator. The subjects could decide on inventory control variables and the amount and type of information to be used in monitoring system performance. The demand for inventory was an external random variable under the control of the experimenter. The experiment investigated how different demand variances affected the decision information used, the decisions made, and the resulting decision effectiveness (cost.) It was found that decision review frequency was not affected by demand variability. However, age and degree of summary of information used were greatly affected. Subjects exposed to high variability used data with a shorter history and a higher level of detail than those exposed to low variability. It was also found that the number of reports used increased from the low to middle variance group, then decreased at very high variance. Subjects tended to “give-up” on their information system at high variance, and they relied on additional safety stock to prevent frequent stockouts. Finally, the correlation between the information used and decision effectiveness as measured by cost was low. This result was contrary to the subjects' beliefs that more and better information produced “better” decisions. It indicates that although variability may strongly affect preferences for different types of information, the information used may not in turn affect decision performance.     

AN ADAPTIVE MULTIVARIATE APPROACH TO TIME SERIES FORECASTING

In recent years, time series analysts have shifted their interest from univariate to multivariate forecasting approaches. Among them, the Box-Jenkins transfer function process and the state space method have received the most attention. This paper presents a simplified approach that embodies some desirable features of existing methods. It stresses empirical analysis, has a unified modeling structure, is easily applicable, and is adaptive to changes without necessitating prior information on the evolution of a system under study. The core of the method relies on the Carbone-Longini adaptive estimation procedure (AEP). Results of a comparative study based on the well-known Lydia E. Pinkham data and the Box-Jenkins sales/leading indicator data illustrate the merits of multivariate AEP in improving forecasting accuracy while simplifying the analysis process. Subject Area: Forecasting.     

ADVERTISING AND VALUE MAXIMIZATION: AN INTEGRATIVE APPROACH *

The model developed in this paper is interdisciplinary in that we incorporate the advertising decision into the capital asset pricing model (CAPM) for valuation of risky income streams. The advertising decision involves the purchase of media services whose effects on sales (profit) are hypothesized to vary with general economic activity. As in any asset expansion, the firm purchases an income distribution with a given expected value and covariance with the economy. By consulting the CAPM, we are able to obtain a value for the income distribution associated with the advertising expenditure. By failing to account for both risk and return, those previous studies which have examined the effect of advertising on profit or valuation are, at best, incomplete. We are able to demonstrate the inappropriateness of ranking alternative advertising strategies solely on the basis of expected income.     

FOUR APPROACHES TO THE CLASSIFICATION PROBLEM IN DISCRIMINANT ANALYSIS: AN EXPERIMENTAL STUDY*

Four discriminant models were compared in a simulation study: Fisher's linear discriminant function [14], Smith's quadratic discriminant function [34], the logistic discriminant model, and a model based on linear programming [17]. The study was conducted to estimate expected rates of misclassification for these four procedures when observations were sampled from a variety of normal and nonnormal distributions. In contrast to previous research, data were taken from four types of Kurtotic population distributions. The results indicate the four discriminant procedures are robust toward data from many types of distributions. The misclassification rates for both the logistic discriminant model and the formulation based on linear programming consistently decreased as the kurtosis in the data increased. The decreases, however, were of small magnitude. None of these procedures yielded statistically significant lower rates of misclassification under nonnormality. The quadratic discriminant function produced significantly lower error rates when the variances across groups were heterogeneous.     

A BAYESIAN APPROACH TO DISCRIMINATION AMONG ECONOMIC MODELS*

Frequently, alternative theories are proposed as relevant models for an economic process. Which alternative is a relatively better representation is a question of fact, but empirical methods of resolving the issue have not been compelling. This study developes a technique for discrimination from Bayesian statistical theory. This technique can analyze an arbitrary number of alternatives simultaneously, and each alternative can be characterized by distribution theory appropriate to its structure. An example, the aggregate consumption function, is considered, and each alternative is characterized by the Normal regression process. The results may only be regarded as suggestive, as the parameterization of the structure may be incomplete on that assumption.