Helen M. Moshkovich

Verbal decision analysis for unstructured problems

Introduction. 1. Individual Decision Making: Approaches and Methods. 2. A New Approach to Unstructured Problems of Decision Making. 3. The Method ZAPROS-LM For Partial Rank-Ordering of Multiattribute Alternatives. 4. The Method PACOM for the Selection of the Best Alternative. 5. The Method ORCLASS For Ordinal Classification of Multiattribute Alternatives. Conclusion. References. Index.

Data influences the result more than preferences: some lessons from implementation of multiattribute techniques in a real decision task

Abstract The concept of model management as a logical extension and parallel of data management has resulted in attempts to extend existing data models to incorporate modeling functionality. This fosters a data-oriented view which artificially restricts the domain of model management. This paper suggests that a model-oriented approach which views ‘data as a model’ not only expands the scope of model management but also offers a more integrated and balanced conceptual foundation from which to implement model management. On the organizational level, this approach recognizes the transition from informal modeling as exemplified by the use of spreadsheets to formal modeling which manages models as a resource in conjunction with data. The notion of information administration is introduced as an organizational mechanism for controlling this evolution. On the technological level, a generalized model management system (GMMS) is required for support of organizational modeling activities. Geoffrions structured modeling is suggested as a foundation from which to build a model-oriented GMMS. The incorporation of artificial intelligence capabilities into a GMMS is seen as a second implementation step which require the development of a sufficiently flexible meta-level architecture. Model abstractions are introduced as a vehicle for implementing this architecture.

ZAPROS-LM — A method and system for ordering multiattribute alternatives

Abstract A method to aid in qualitative evaluation of multiattribute alternatives is proposed. It not only elicits information from a decision-maker in a qualitative form but tries to use it without resort to numbers, and to apply rational logic for comparison of alternatives. Special procedures for identification of possible inconsistencies in decision-makers information and elimination of them in a dialogue with a decision-maker are developed. Possibilities for verification and explanation of the results for partial ordering of a large set of alternatives are shown. Two main assumptions are used: transitivity of the decision-makers preferences and preferential independence of attributes. Problems of justification of these properties in real tasks of decision making are discussed. The description is accompanied by an example.

Ordinal judgments in multiattribute decision analysis

Abstract The article discusses the contradiction between the ambiguity of human judgment in a multicriterion environment and the exactness of the assessments required in the majority of the decision-making methods. Preferential information from the decision makers in the ordinal form (e.g., “more preferable”, “less preferable”, etc.) is argued to be more stable and more reliable than cardinal input. Ways of obtaining and using ordinal judgments for rank ordering of multiattribute alternatives are discussed. The effectiveness of the step-wise procedure of using ordinal tradeoffs for comparison of alternatives is evaluated. We introduce the notion of ordinal tradeoffs, presentation of ordinal tradeoffs as a flexible three-stage process, a paired joint ordinal scale (PJOS), and evaluation of the effectiveness of the three-stage process. Simulation results examine the sensitivity of the number of pairwise comparisons required for given numbers of criteria and categories within criteria, as well as the number of alternatives analyzed. This simulation shows that ordinal pairwise comparisons provide sufficient power to discriminate between 75% and 80% of the alternatives compared. While the proportional number of pairwise comparisons relative to the maximum possible decreases with the number of criteria and categories, the method is relatively insensitive to the number of alternatives considered.

Marketing decisions in small businesses: how verbal decision analysis can help

Marketing decisions are crucial in the strategic planning and are usually a starting point for the analysis. Marketing practices of small businesses are different from those of large companies and require different approaches to decision support. The paper illustrates the use of the ORCLASS method of verbal decision analysis in a real application. The ORCLASS method is based solely on the use of qualitative (verbal) information and allows the process leading to decisions to be transparent and shared with the stakeholders in a clear way. The research results show that, the adoption of a qualitative method of decision support in particular, can benefit significantly the marketing of a small business.

Rule induction in data mining: effect of ordinal scales

Abstract Many classification tasks can be viewed as ordinal. Use of numeric information usually provides possibilities for more powerful analysis than ordinal data. On the other hand, ordinal data allows more powerful analysis when compared to nominal data. It is therefore important not to overlook knowledge about ordinal dependencies in data sets used in data mining. This paper investigates data mining support available from ordinal data. The effect of considering ordinal dependencies in the data set on the overall results of constructing decision trees and induction rules is illustrated. The degree of improved prediction of ordinal over nominal data is demonstrated. When data was very representative and consistent, use of ordinal information reduced the number of final rules with a lower error rate. Data treatment alternatives are presented to deal with data sets having greater imperfections.

Verbal Decision Analysis: Foundations and Trends

The primary goal of research in multiple criteria decision analysis is to develop tools to help people make more reasonable decisions. In many cases, the development of such tools requires the combination of knowledge derived from such areas as applied mathematics, cognitive psychology, and organizational behavior. Verbal Decision Analysis (VDA) is an example of such a combination. It is based on valid mathematical principles, takes into account peculiarities of human information processing system, and fits the decision process into existing organizational environments. The basic underpinnings of Verbal Decision Analysis are demonstrated by early VDA methods, such as ZAPROS and ORCLASS. New trends in their later modifications are discussed. Published applications of VDA methods are presented to support the findings.

Comparison of MCDA Paradigms

The underlying concepts of MAUT, SMART, AHP, preference cones, ZAPROS, and outranking methods are compared. Learning systems are considered. The learning view is that decision makers initially do not fully understand all of the criteria that are important. Therefore, rather than uncovering an underlying utility function, what must be uncovered are the full ramifications involved in selecting one alternative over another. This paradigm can involve an evolutionary problem, where criteria can be added or discarded during the analysis. Methods are also reviewed with respect to their psychological validity in generating input data. Past experiments conducted by the authors are reviewed, with conclusions drawn relative to subject comfort in using each method. Subjects typically make errors, in that they have inconsistent ratings of scores across systems, and will occasionally have reversal of relative importance of criteria across systems. This emphasizes the need to be careful of input in decision models, and strengthens the argument for more robust input information. Furthermore, systems based on the same model have been found to yield different results for some. In a study exposing both US and Russian students were compared. Each group found it more comfortable to use systems developed within their own culture.

Cognitive Effort and Learning Features of Decision Aids: Review of Experiments

ABSTRACT Decision aids are computer systems intended to help decision makers select from a set of alternatives when considering multiple criteria. There are a number of basic ideas behind such systems, including multiattribute utility analysis, analytic hierarchy process, and outranking. Some systems focused on identification of decision maker preference functions. Recently there have been a number of systems focusing on the idea of enhancing decision maker learning. This paper reviews the fundamental ideas of a number of decision aids and considers the cognitive effort on the part of decision makers required to use each system. Recent empirical studies from a number of sources are reviewed and evaluated for support of theories about this cognitive effort, as well as how learning systems are implemented.

Russian Business Schools in a Time of Transition.

In this study, the authors reviewed the development of Russian business education in the past decade. This development, fueled by historic changes in Russian society, has affected all aspects of business education, including its organizational structures, demand in different business areas, and mode of teaching. In a short period of time, Russian business education has substantially adjusted to new market requirements and transformed itself into one of the most influential parts of the Russian academic community. However, business education in Russia still has many challenging problems inherited from the Soviet-era planning mentality, which emphasized theory at the expense of practical problem solving.