Selecting Attributes to Measure the Achievement of Objectives

Authors:
Ralph L. Keeney;Robin S. Gregory
Affiliations:
-;-
Venue:
Operations Research
Year:
2005

Citing 0
Cited 6

Evaluating Academic Programs: With Applications to U.S. Graduate Decision Science Programs

Operations Research
Smart-Swaps - A decision support system for multicriteria decision analysis with the even swaps method

Decision Support Systems
Studying transition dynamics via focusing on underlying feedback interactions

Computational & Mathematical Organization Theory
Modeling challenges with influence diagrams: Constructing probability and utility models

Decision Support Systems
One-Switch Independence for Multiattribute Utility Functions

Operations Research
Toward an Improved Methodology to Construct and Reconcile Decision Analytic Preference Judgments

Decision Analysis

Quantified Score

Hi-index	0.01

Visualization

Abstract

The foundation for any decision is a clear statement of objectives. Attributes clarify the meaning of each objective and are required to measure the consequences of different alternatives. Unfortunately, insufficient thought typically is given to the choice of attributes. This paper addresses this problem by presenting theory and guidelines for identifying appropriate attributes. We define five desirable properties of attributes: they should be unambiguous, comprehensive, direct, operational, and understandable. Each of these properties is discussed and illustrated with examples, including several cases in which one or more of the desirable properties are not met. We also present a decision model for selecting among the different types of natural, proxy, and constructed attributes.