Decision Analysis Applications in the Operations Research Literature, 1970-1989

Author(s): James L. Corner and Craig W. Kirkwood

Source: Operations Research , Mar. - Apr., 1991, Vol. 39, No. 2 (Mar. - Apr., 1991), pp.
206-219
Published by: INFORMS

Stable URL: https://www.jstor.org/stable/171098

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/171098?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

INFORMS is collaborating with JSTOR to digitize, preserve and extend access to Operations
Research

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
 DECISION ANALYSIS APPLICATIONS IN THE OPERATIONS
RESEARCH LITERATURE, 1970-1989

JAMES L. CORNER AND CRAIG W. KIRKWOOD

Arizona State University, Tempe, Arizona

(Received September 1989; revision received May 1990; accepted July 1990)

In order to provide a guide to source material for practitioners interested in applying decision analysis methods, this paper
surveys applications of decision analysis published from 1970 through 1989. In addition, it presents references for useful
decision analysis methods that are often omitted from introductory textbooks. As used in this article, the term decision analysis
refers to a set of quantitative methods for analyzing decisions which use expected utility as the criterion for identifying the
preferred decision alternative. To be included in this survey, an application had to explicitly analyze alternatives for a decision
problem using judgmental probabilities and/or subjectively assessed utility functions. The paper classifies the applications into
five areas: energy, manufacturing and services, medical, public policy, and general. It further subclassifies energy applications
into bidding, product and project selection, regulation, site selection, and technology choice. Those in manufacturing and
services are subclassified into budget allocation, product planning, strategy, and miscellaneous. Applications in public policy are
subclassified into standard-setting and miscellaneous. The paper notes articles that present significant detail about methodologi-
cal and implementation issues, including problem structure/formulation, decision trees, probability and utility assessment,
communication/facilitation, and group decision making.

T his article surveys applications of decision analysis long time-horizons. As the applications show, decision
that appeared in major English language operations analysis provides approaches to address these issues
research journals and other closely related journals from quantitatively in a practical, as well as logically defensi-
1970 through 1989. These are classified by application ble, manner. We emphasize this point because the recent
area, as well as by methodological or implementation report by the prestigious Committee On The Next Decade
issue. In addition, references are presented for useful in Operations Research (1988), as well as follow-up
decision analysis methods that are often not covered in comments (Wagner et al. 1989), do not mention decision
introductory textbooks. The intent is to provide a guide analysis and barely mention the type of strategy/policy
to relevant source material for operations research practi- questions considered by many applications surveyed in
tioners facing a situation where decision analysis might this article. This leaves the impression that operations
potentially be applicable. Decision analysis provides tools research methods can only address repetitive, opera-
for quantitatively analyzing decisions with uncertainty tional/tactical problems. In fact, this is not true as the
and/or multiple conflicting objectives. These tools are applications surveyed below make clear.
especially useful when there is limited directly relevant Section 1 defines decision analysis and provides refer-
data so that expert judgment plays a significant role in ences to further information on basic decision analysis
the decision making process. Such situations include methods. Section 2 surveys 86 decision analysis applica-
government policy making and regulation, strategic busi- tions by application area, while Section 3 summarizes
ness decisions, and such risky personal decisions as the methodological and implementation issues empha-
selecting a treatment for a serious medical problem. sized by each application. Table I lists the articles by
It is of particular note that many decision analysis application area, while Table II lists them by method-
applications address decisions with strategic or policy ological and/or implementation issue.
implications. These are generally characterized by one
or more of the following characteristics: multiple con-
1. DECISION ANALYSIS
flicting objectives, limited directly relevant data, multi-
ple interested stakeholders, alternatives that differ from As used in this article, the term decision analysis refers
each other qualitatively as well as quantitatively, uncer- to a set of quantitative methods for analyzing decisions
tainties that pose significant organizational risks, and based on the "axioms of consistent choice" (Pratt,

Subject classifications: Decision analysis, applications: survey of applications. Utility/preferences, applications: survey of applications.

Operations Research 0030-364X/9 1/3902-0206 $01.25

Vol. 39, No. 2, March-April 1991 206 ?~ 1991 Operations Research Society of America

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
 Decision Analysis Applications Survey / 207

Raiffa, and Schlaifer 1964, Keeney 1982). These meth- Spetzler and Stael von Holstein (1975) and Merkhofer
ods use expected utility as the criterion for identifying (1987) summarize probability assessment methods.
the preferred decision alternative, and they guarantee For decisions where attitude toward risk taking or
that complex decisions will avoid such undesirable char- tradeoffs among conflicting objectives are important,
acteristics as intransitivity of preferences (i.e., willing- utility functions are used to analyze these aspects of the
ness to serve as a "money pump") (von Winterfeldt and decision. Howard (1988) notes that in situations with a
Edwards 1986). Decision analysis is normative, rather single evaluation measure the exact form of the utility
than descriptive. That is, it provides a systematic quanti- function often does not change the decision. He also
tative approach to making better decisions, rather than a shows some evidence of a systematic relationship be-
description of how unaided decisions are made. As tween company size and attitude toward risk taking. In
Howard (1988) notes, the focus is on "transforming situations with multiple objectives, sensitivity analysis
opaque decision problems into transparent decision prob- often shows that only a few tradeoffs affect the final
lems by t sequence of transparent steps," and the ap- decision. Keeney and Raiffa (1976) and Keeney (1977)
proach "offers the possibility to a decision-maker of discuss development of a multiattribute utility function in
replacing confusion by clear insight into a desired course detail.
of action." The level of detail for the models used in decision
Decision analysis applications begin with the determi- analysis applications varies. In some studies, highly de-
nation of the decision alternatives to be analyzed and tailed structural models (sometimes referred to as sys-
specification of the evaluation measures (attributes) to be tem or process models) are developed for the decision
used to measure the degree of attainment of each deci- situation, while in others less emphasis is placed on this.
sion objective. Howard (1988), Keller and Ho (1988) The structural models for many decisions must account
and McNamee and Celona (1987) discuss the develop- for possible reactions of other stakeholders to the deci-
ment of alternatives, and Keeney (1988) and Keeney and sion that is made. Such methods as discrete choice
Raiffa (1976) discuss the determination of evaluation analysis (Amemiya 1981, Ben-Akiva and Lerman 1985)
measures. The specification of alternatives and evalua- can be used to model these reactions.
tion measures is usually followed by construction of an
initial model of the performance of each decision alterna- 2. SURVEY OF DECISION ANALYSIS
tive with regard to each evaluation measure. Influence APPLICATIONS
diagrams can aid in developing such models (Howard
The following journals were exhaustively reviewed from
1988, Kirkwood 1990, Howard and Matheson 1984a,
1970 through 1989 to determine decision analysis appli-
McNamee and Celona 1987). Once the initial model is
cation articles (the numbers in parentheses are the num-
constructed, sensitivity analysis is often conducted to
ber of application articles identified in each journal):
determine which uncertainties potentially affect the deci-
* Decision Sciences (4)
sion. Tornado diagrams can be useful for presenting the
* European Journal of Operational Research (5)
results of the sensitivity analysis (Howard 1988,
* IEEE Transactions on Systems, Man, and Cyber-
McNamee and Celona 1987).
netics (formerly the IEEE Transactions on Systems
Decision trees (Holloway 1979, Samson 1988) can
Science and Cybernetics) (7)
provide a useful modeling approach for the uncertainties
* Interfaces (18)
in a decision. For realistic decision situations, decision
* Journal of the Operational Research Society
trees are often too large for hand construction or analy-
(formerly Operational Research Quarterly) (18)
sis. Specialized software is available to assist with build-
* Management Science (10)
ing and analyzing tree models (Celona and McNamee
* Omega (3)
1988, Cheung and Kirkwood 1989, Texas Instruments,
* Operations Research (15)
Inc. 1985). Decision tree models of situations with con-
* Operations Research Letters (0), and
tinuous random variables may require development of
* Risk Analysis (6).
discrete approximations for the continuous probability
distributions. Keefer and Bodily (1983) and Miller and The identified articles are classified into these applica-
Rice (1983) review practical and accurate approximation tions areas and subareas:

methods. In situations where expert judgment is used to * Energy

determine probability distributions, care must be taken to Bidding
avoid the well documented difficulties that people have Product and Project Selection
in accurately expressing their uncertainty (Dawes 1988). Regulation

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
208 / CORNER AND KIRKWOOD

Site Selection plications prior to 1980. Keeney and Raiffa (1976) and
Technology Choice Bell, Keeney and Raiffa (1977) include a number of
* Manufacturing and Services applications with multiple objectives. French (1989)
Budget Allocation reprints several applications that originally appeared in
Product Planning the Journal of the Operational Research Society,
Strategy including one prior to the period covered in this survey.
Miscellaneous Howard and Matheson (1984b) include applications to
* Medical
investment and strategic planning, research and develop-
* Public Policy
ment, and social policy. Ulvila and Brown (1982)
Standard Setting
summarize other applications.
Miscellaneous
Some decision analysis applications appear only in
* General
consultant or government reports. While many of these
Those application articles presenting significant detail are proprietary, two organizations have sponsored sub-
about a particular decision analysis methodological or stantial decision analysis work for which reports are
implementation issue are noted in Section 3. available. The Electric Power Research Institute has
There is some subjectivity in deciding whether a par- funded decision analysis applications in areas of interest
ticular application is actually decision analysis. To be to the electric power industry, and reports on this work
included here, an application had to explicitly analyze are available. Contact the Technical Information Center,
alternatives for a decision problem using judgmental Electric Power Research Institute, 3412 Hillview
probabilities and/or subjectively assessed utility func- Avenue, P. 0. Box 10412, Palo Alto, California 94303
tions. When there is ambiguity about whether an article (telephone: (415) 855-2411). The National Technical
meets this requirement, this usually revolves around: Information Service (NTIS) of the U.S. Department of
1) whether probabilities were judgmentally assessed, or Commerce indexes reports for decision analysis work
2) whether conceptually correct methods were used to conducted by or for U.S. federal agencies. The NTIS
determine a utility function over multiple evaluation issues a Government Reports Annual Index, and these
measures. Ambiguous cases were resolved by including are available at research libraries.
the article if, on balance, it took a decision analysis
Energy
approach.
There is also some subjectivity in deciding whether an Energy applications are subcategorized into bidding,

article reports an application. A majority of the surveyed product and project selection, regulation, site selection,
articles report case histories of the use of decision analy- and technology choice.

sis to address a specific decision problem. Other articles Bidding. Bell (1984) analyzes a coal company's selec-
report on analysis performed to provide background for tion of a method to haul coal. The alternatives consid-
policy making. In a few cases, there is no application but ered are making a bid to salvage a grounded ship,
the material is of direct interest for applications. purchase of a new ship, or subcontracting for delivery.
The application areas are considered in alphabetical A decision tree is constructed, and expected net present
order with the exception that the General category is value is used as the evaluation measure to determine the
discussed last. Under each application area, the subareas preferred bid price for the grounded ship. The analysis
are discussed in alphabetical order with the exception considers potential bids by a competitor.
that the Miscellaneous subarea is discussed last for areas Dyer and Lorber (1982) consider the evaluation of
containing this subarea. Within each subarea, articles are potential subcontractors. Eleven evaluation criteria are
considered in alphabetical order by authors' last names identified for use in evaluating three subcontractor pro-
with the exception of articles addressing the same appli- posals to conduct a program planning project. Since each
cation and the medical applications. (The nature of the subcontractor had completed a test project before the
health-related issues considered by the medical articles final proposal evaluation was conducted, there was little
leads naturally to cross-categorizing these as shown uncertainty about subcontractor performance. Thus, un-
below.) The articles are listed by application area in certainty is not considered in the final evaluation. A
Table I. measurable value function is used to combine the evalua-
The surveyed journals have published a wide variety tion criteria. The analysis shows that varying weights for
of decision analysis applications, but applications also evaluation measures does not affect the final evaluation
appear in other places. Keeney (1982) references several results. Sarin, Sicherman and Nair (1978) consider pro-
applications that have appeared outside the operations posal evaluation in the context of large-scale solar total
research literature. Krischer (1980) reviews medical ap- energy experiments. Seventeen evaluation measures are

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
 Decision Analysis Applications Survey / 209

Table I
Articles Listed By Application Area

Energy

Bidding: Bell (1984), Dyer and Lorber (1982), Sarin, Sicherman and Nair (1978).
Product and Project Selection: Crawford, Huntzinger and Kirkwood (1978), Kirkwood and Sarin (1985), Madden, Hyrnick and Hodde
(1983), Peerenboom, Buehring and Joseph (1989).
Regulation: Judd and Weissenberger (1982), Keeney and Smith (1982), Lathrop and Watson (1982), Lincoln and Rubin (1979), von
Winterfeldt (1982).
Site Selection: Allett (1986), Gregory and Lichtenstein (1987), Hobbs (1980), Hosseini (1986), Keeney (1979), Keeney (1987), Kirkwood
(1982), Merkhofer and Keeney (1987).
Technology Choice: Golabi, Kirkwood and Sicherman (1981), Keeney and Sicherman (1983), Keeney, Lathrop and Sicherman (1986), North
and Stengel (1982).

Manufacturing and Services

Budget Allocation: Brooks and Kirkwood (1988), Keefer and Kirkwood (1978), Ozernoy, Smith and Sicherman (1981).
Product Planning: Longbottom (1973), Phillips (1982), Ronen and Pliskin (1981), Smallwood and Morris (1980).
Strategy: Digman (1980), Dyer and Lund (1982), Higgins (1982), Lock (1982), Ulvila, Brown and Packard (1977).
Miscellaneous: Belton (1985), Wells (1982), Winter (1985), Wooler (1982).

Medical

Alemi and Angliato (1989), Betaque and Gorry (1971), Bodily (1977), Clarke (1987), Forst (1974), Fryback and Keeney (1983), Mehrez and
Gafni (1987), Pliskin, Ronen and Feldman (1987), Pliskin, Shepard and Weinstein (1980), Ronen, Pliskin and Feldman (1984), Sanchez de
Rivera (1980), Shachtman (1980), Torrance, Boyle and Horwood (1982), Venta and Venta (1987), Zalkind and Shachtman (1983)

Public Policy

Standard Setting: Anandalingam (1987), Anandalingam (1989), Edwards (1977), Harvey (1983), Jensen, Tome and Darby (1989), Keeney
and Ozernoy (1982), Ulvila and Snider (1980), van Steen (1987).
Miscellaneous: Anandalingam and Olsson (1989), Bodily (1978), Chen et al. (1979), Cohan et al. (1984), Edwards and von Winterfeldt
(1987), Manne, Richels and Weyant (1979), Peck (1980), Sarin (1983), Ulvila (1987), Ulvila (1988), Ulvila and Seaver (1982), Wenstop
and Carlsen (1988).

General

Dalkey (1981), Dyer and Miles (1976), Farmer (1987), Heian and Gale (1988), Janssen and Daniel (1984), Longbottom and Wade (1973),
Luna and Reid (1986), Pearman (1987), Pollock and Chen (1986).

combined using a multiplicative utility function. Eight tradeoffs among the various alternatives. The analysis
evaluators assigned values to the evaluation measures for identifies a subset of the borehole plugging materials to
each proposal. be given more detailed study.
Madden, Hyrnick and Hodde (1983) discuss the choice
Product and Project Selection. Crawford, Huntzinger
between fabric filters and electrostatic precipitators for
and Kirkwood (1978) conduct a decision analysis of a controlling sulfur dioxide emissions from a coal-fired
power company's choice from among several
electric power generating plant. The final decision was
tower/conductor alternatives for a proposed 765 KV made by comparing the cumulative probability distribu-
electric transmission line. A deterministic analysis is tion functions for each product choice. Peerenboom,
first conducted, and a sensitivity analysis on uncertain Buehring and Joseph (1989) describe the use of a hierar-
elements shows that the uncertainty might affect the chical decision analysis procedure to develop a portfolio
preferred alternative. Thus, a complete probabilistic of environmental and health research programs for a
analysis is done. There are three financial evaluation commercial-scale synthetic fuels facility. The overall
measures and one environmental evaluation measure. areas of concern in evaluating the 88 proposed projects
Kirkwood and Sarin (1985) consider selection of bore- were project comprehensiveness, relevance, and cost
hole plugging materials for underground radioactive effectiveness.
waste storage. They develop a pairwise comparison pro-
cedure for ranking alternatives when only partial infor- Regulation. A complication of regulatory decision anal-
mation is known about the weighting constants for the ysis is deciding whose preferences or values should be
various evaluation measures. This procedure determines used in the decision making. The articles on energy
the degree of ranking of the various alternatives that is regulatory decision making discussed below address pol-
possible when there is only partial information about the icy issues related to regulatory actions and illustrate the

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
210 / CORNER AND KIRKWOOD

use of decision analysis methods for this type of deci- in the final decision. After brief reviews of different
sion. Judd and Weissenberger (1982) consider the prob- methods for selecting these weights, two methods are
lem of regulating nuclear material safeguards. Costs and compared using a set of 18 evaluation measures deter-
uncertainties associated with various nuclear material mined by five siting experts. Significantly different
control and accounting systems are considered. In addi- decisions are made using the two methods.
tion, potential adversaries of such systems are modeled Hosseini (1986) considers selection of a location for a
as well as the consequences of various adversary actions. wildcat oil well. Considerable detail is provided on the
The analysis of the resulting decision tree model shows analysis of geological and engineering uncertainties, as
that selecting the optimal safeguards system requires well as those due to economic and political factors.
consideration of the tradeoffs between the cost of the Judgmental probabilities are used to analyze these uncer-
system and the degree of risk of material diversion. tainties, and a decision tree is used to evaluate the net
Lathrop and Watson (1982) apply an expected utility present value of two potential well sites based on the
approach to develop risk indices for uncertain health- assessed probabilities as well as the producer's attitude
effect consequences of a nuclear waste management sys- toward risk taking.
tem. Preferences of different interested groups are com- Keeney (1979) presents a multiattribute decision anal-
bined into an overall group function with this approach ysis to select a site for a pumped storage facility in the
in order to set regulations, but the method has not Southwest United States. Evaluation measures for cost,
actually been used in setting regulations. Lincoln and transmission line distance, and two environmental impact
Rubin (1979) consider cross-media environmental regu- considerations are used to assess ten candidate sites.
lation. This arises when regulations to reduce one pollu- Preferences of key company personnel are obtained.
tion type (e.g., air) may lead to increases in other types Kirkwood (1982) summarizes a siting study for a 2,500
of pollution (e.g., water or land). Thus, tradeoffs must megawatt nuclear power plant in the western United
be made between several interdependent emissions. The States. An initial screening process is used to eliminate
authors illustrate how multiattribute utility theory could areas with obvious undesirable characteristics. In this
be applied to this problem. The results for an illustrative study, the screening process is decoupled into a search
coal-fired generation scenario show that the key consid- for sources with adequate water availability and a search
eration is the tradeoff between solid waste and SO2 for plant sites that meet certain requirements on health
emissions. The implication is that no single regulation or and safety, environmental impact, and systems cost and
technology is optimal with respect to all pollutants. reliability. Candidate water sources and plant sites are
Another regulatory model is found in Keeney and then recoupled for evaluation by a multiattribute decision
Smith (1982). This article examines the impact on nu- analysis process involving nine evaluation measures.
clear facility operators of regulations for nuclear mate- Merkhofer and Keeney (1987) present a multiattribute
rial control and accounting. Evaluation considerations decision analysis conducted to aid the U.S. Department
include employer impact, implementation costs, and reg- of Energy in its selection of three locations to study in
ulatory uncertainty. The article shows the usefulness more detail as potential sites for a mined geologic reposi-
of decision analysis methods for incorporating values tory for nuclear waste. Five locations that had previously
into the analysis of a complex policy problem. Von been recommended are evaluated with regard to health
Winterfeldt (1982) considers regulation of chronic off- and safety, environmental, socioeconomic, and cost fac-
shore platform oil discharge in the United Kingdom. tors. The article discusses the role of this decision analy-
Three separate models are developed for governmental, sis in the decision making process of the Department of
environmental, and operator stakeholders in the regula- Energy. Gregory and Lichtenstein (1987) review the
tory process. These models are used to evaluate different Merkhofer and Keeney study as well as another report
regulatory procedures for chronic offshore oil discharge by the Department of Energy on repository site
from the standpoints of the three interested parties. selection. Keeney (1987) extends the work in Merkhofer
and Keeney (1987) to explicitly consider portfolio as-
Site Selection. Allett (1986) presents an illustration of pects of the selection of three sites. He considers possi-
how decision analysis might be applied to coal mine ble probabilistic interdependencies among the sites, as
siting. This article presents less detail than the other well as sequential study strategies for the sites.
siting studies discussed below, but a strength is its
comparison of multiattribute decision analysis with cost Technology Choice. Both energy producers and energy
benefit analysis for the analysis of environmental im- regulators have applied decision analysis to ascertain
pacts. Hobbs (1980) discusses a nuclear power plant proper routes of development for energy-related
siting problem in the state of Maryland. The focus is to technologies. Golabi, Kirkwood and Sicherman (1981)
investigate the importance of evaluation measure weights combine decision analysis and optimization methods to

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
 Decision Analysis Applications Survey / 211
evaluate solar energy project proposals. There are no used to reduce the initial list of 92 potential systems to a
significant uncertainities because the technologies in- final list of five candidates. A multiattribute utility func-
volved are proven. The goal of the selection process is to tion is developed to rank the final candidates.
choose an optimal combination of projects to meet U.S.
Department of Energy objectives. Technical quality and Product Planning. Longbottom (1973) considers the
two programmatic considerations (budget and diversity) analysis of a refrigeration firm's future resource require-
are used as measures of project acceptability. A measur- ments for two noncompeting refrigerated container
able value function approach is used to develop an product lines. The author notes the importance of
overall measure of the technical quality of the projects. good communication and company commitment for the
Programmatic considerations are incorporated as con- acceptance of the analysis. Phillips (1982) presents a
straints in a 0-1 linear integer programming optimization "requisite" decision modeling approach for deciding
approach to determine the final set of projects to be whether to continue producing an existing marine prod-
funded. uct which faces a possible future commercial ban, or to
Studies performed for specific electric utilities are redesign the product. The requisite modeling approach
found in Keeney and Sicherman (1983) and Keeney, emphasizes an iterative process, including group discus-
Lathrop and Sicherman (1986). Both studies utilize the sions, to ensure that no part of the problem is omitted
Electric Power Research Institute's technology choice from the analysis. The article includes a discussion
model as an analysis tool. Keeney and Sicherman con- of typical pitfalls encountered in decision analysis
sider the choice between coal and nuclear generating applications.
technologies, while Keeney, Lathrop and Sicherman Ronen and Pliskin (1981) consider a system reliability
consider the choice between selecting current coal gener- decision in the Israeli microelectronics industry. The
ating technology or waiting for development of advanced decision involves whether or not to include redundancy
technology. in a diode array circuit as well as the choice of a specific
North and Stengel (1982) use a decision tree model packaging technology. Two analyses are conducted, one
with a single evaluation measure (benefits minus costs) considering only cost, and one considering three evalua-
to analyze the choice between the alternatives of toka- tion concerns (cost, damage, and flexibility). Smallwood
mak and magnetic fusion test facilities. The extent of and Morris (1980) consider the integrating role of the
uncertainties regarding fusion development is reflected analyst in interorganizational decision making. Their
by the 11,000 end-points on the decision tree. analysis considers the potential need for increased manu-
facturing capacity to produce consumables used by a new
Manufacturing and Services photocopier. A decision tree model is used, with various
The manufacturing and services applications are classi- submodels providing information for the tree.
fied into budget allocation, product planning, strategy,
and miscellaneous. Strategy. Digman (1980) presents a simplified analysis
of strategies to regain market share in the airline industry
Budget Allocation. Brooks and Kirkwood (1988) pre- following settlement of a strike. In addition to decision
sent a multiattribute decision analysis of a public utility's analysis, such approaches as game theory and Markov
strategies for dealing with local area networking to link models are considered for analyzing this problem. The
their microcomputers. The existence of multiple inter- application is illustrative of how to form a real-world
ested groups and little hard technical data or expertise decision problem into a classroom case discussion. Dyer
compound the problem. A five-step methodology is used and Lund (1982) analyze new strategies for merchandis-
which considers nine evaluation measures. ing gasoline and other products through full-facility ser-
Keefer and Kirkwood (1978) consider the allocation of vice stations. The decision analysis led to a significant
a department's annual product engineering budget. Vari- change in marketing strategy which had a large impact
ous organizational pressures as well as departmental on sales.
budgetary limitations are considered. The lack of natural Higgins (1982) presents two case studies; one involv-
evaluation measures means that special evaluation scales ing pricing of advertising space by a newspaper pub-
need to be constructed for some of the evaluation consid- lisher in light of anticipated competitors' actions, and
erations. A multiattribute utility function is determined one involving a litigation decision by a food retailer.
for the department director, and the final decision model Both are board level decisions. While the article does not
is a nonlinear optimization model. Ozernoy, Smith and contain much detail about the specific analyses, there is a
Sicherman (1981) review a decision analysis application good discussion of the use of decision analysis at the
to select a computerized geographic information system upper levels of management. Lock (1982) presents an
for a professional consulting firm. A screening process is analysis of the potential need to increase a mens shoe

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
212 / CORNER AND KIRKWOOD

firm's business activity. Eight alternative strategies and and Feldman 1987, and Ronen, Pliskin and Feldman
twelve evaluation measures are considered. 1984) to diagnosis of disease (Betaque and Gorry 1971,
Ulvila, Brown and Packard (1977) analyze a firm's Sanchez de Rivera 1980, and Venta and Venta 1987).
opportunity to invest in the patent rights for a flight These applications can also be classified by whether they
safety product. Uncertainties about product acceptability, deal directly with patients and illness (Clarke 1987,
government regulatory rulings, and potential defense Fryback and Keeney 1983, Mehrez and Gafni 1987,
contract awards are considered in a decision tree model. Sanchez de Rivera 1980, Venta and Venta 1987, and
Judgmental probabilities are obtained by decomposing Zalkind and Shachtman 1983) or with the physician (the
uncertain quantities and assessing conditional distribu- remainder of the applications considered in this section).
tions. Review of the decision tree model leads to devel- Before considering the applications articles in more
opment of additional decision alternatives. detail, note that four articles provide general background
on the use of decision analysis in medical decision
Miscellaneous. Belton (1985) uses a multiattribute value making. Forst (1974) reviews four of the conditions
function approach to assist in selecting, from a short-list needed to assume the existence of a decision maker's
of three, the company with which to place a contract for utility function. Pliskin and Pliskin (1980) discuss imple-
the development of a computerized financial manage- mentation issues. They argue that the decision tree
ment system. The article includes a discussion of the analysis structure is attractive to physicians and that
reaction of participants in the analysis process, as deter- judgmental probability assessment provides insight in
mined by a post-analysis questionnaire. situations where data are not available. However, they
One of the remaining three decision analysis applica- say that there are still many problems to overcome to
tions in the manufacturing and service industries deals practically apply decision analysis when the decision
with the general use of decision analysis in a large directly involves a patient. Specifically, they note the
diversified English firm (Wells 1982), while the last two difficulty of determining an appropriate utility function
applications are in the personnel field (Winter 1985 and to use for a patient's decision problem and the impracti-
Wooler 1982). Wells reports on the successful use of cality of determining the patient's utility function in
decision analysis in the Imperial Group since 1975, most cases.
especially the use of decision trees for problem structur- Torrance, Boyle and Horwood (1982) provide a re-
ing and understanding. A specific application to a distri- view of multiattribute utility theory and discuss the con-
bution problem in the food division is considered in ditions necessary for the existence of an additive or
addition to the general discussion. multiplicative utility function. Finally, Venta and Venta
Winter analyzes a heavy industrial manufacturer's la- (1987) discuss the use of decision trees for a decision
bor contract bargaining strategy over multiple time peri- problem with a single evaluation measure, although the
ods. The major uncertainty concerns union decisions to theoretical development is dispersed within the specific
strike or settle. A decision tree approach is coupled with application they consider.
dynamic programming to optimize each strategy for each A number of applications address the patient's utility
possible bargaining period. Winter reports high user function. Clarke (1987) assumes a utility function for the
confidence in the approach due to increased communica- patient in a case involving the decision to operate for
tion and identification of the risks associated with the appendicitis. Here, the condition of the patient precluded
alternative strategies. Wooler discusses two computer- the direct assessment of his/her utility function. Forst
ized decision aids to assist in career choice. His experi- (1974) argues that the patient's utility function should be
ence shows that the decision analysis aids reduce anxi- considered when determining the merits of medical mal-
eties generated by attempting to informally analyze practice cases. Mehrez and Gafni's (1987) study takes
career options with many conflicting objectives. into account the patient's quality and length of life in
deciding on an optimal treatment strategy. Similarly,
Medical
Pliskin, Shepard and Weinstein (1980) develop a utility
Much medical decision analysis work has appeared in function over life years and health status to assist in
the medical decision making literature, which is outside selecting treatments for coronary artery and chronic
the scope of this survey article. Krischer (1980) surveys kidney disease. Sanchez de Rivera (1980) includes length
medical decision analysis applications published between of life as the primary evaluation measure when identify-
the mid-1960s and the late-1970s, mainly in the medical ing disease and prescribing treatment for jaundiced
literature. Studies which have appeared in the operations patients. Torrance, Boyle and Horwood (1982) ag-
research literature range from the decision to operate gregate the utilities of 87 subjects for physical,
(Clarke 1987) to equipment selection (Pliskin, Ronen social activity, emotional, and health attributes into a

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
 Decision Analysis Applications Survey / 213

multiattribute health state classification index. Finally,within a Nash bargaining model. Edwards (1977) consid-
Zalkind and Shachtman (1983) develop a nonmonetary ers several standard setting situations including an appli-
tradeoff method for determining one's personal value for cation of decision analysis for the California Coastal
death in the context of deciding whether to accept a Commission to choose among building development re-
swine influenza vaccination. Although no utility function quests, an evaluation of research recommendations for
is explicitly formulated, values are assessed for tradeoffs the Office of Child Development, and an evaluation of
between probabilities of reaction to the shot, death, and the possibility of having two different water quality
outbreak of an epidemic. standards instead of one. All three of these applications
The remainder of the medical decision analysis appli- include successful resolution of interpersonal disagree-
cations deal with the physician or some other interested ment among multiple expert decision makers using a
party as the decision maker. Alemi and Agliato (1989) multiattribute value assessment procedure.
analyze the decision by a bank to provide incentives to Harvey (1983) considers approaches to a proposed
its employees to use a preferred provider organization regulation by the U.S. Department of Transportation
for health services; Betaque and Gorry (1971) analyze requiring the use of daylight running lights on all new
the diagnosis and treatment of acute renal failure; Bodily highway vehicles. The study does not assign a specific
(1977) considers the optimal utilization level of frozen value per fatality prevented but rather determines ranges
blood in a blood bank; Fryback and Keeney (1983) of values of this quantity for which each alternative is
consider assessment of the trauma severity of injuries; most preferred. Jensen, Tome and Darby (1989) exam-
Pliskin, Ronen and Feldman (1987) and Ronen, Pliskin ine whether residential smoke detectors should be re-
and Feldman (1984) analyze the choice of an appropriate quired by law. Extensive sensitivity analysis is con-
cardiac pacemaker; Shachtman (1980) considers whether ducted to account for deficiencies in available data.
to undertake a national evaluation study on nosocomial Keeney and Ozernoy (1982) analyze the definition of an
infection control; and Venta and Venta (1987) anal- air quality standard for ambient carbon monoxide. The
yze the choice of a diagnostic strategy for deep-vein objective is to develop a standard such that the most
thrombosis. sensitive population group (those with heart related ill-
All the studies that involve probabilities rely on exist- nesses) will not suffer from adverse health effects due to
ing data bases or other readily available published data to the allowed pollution level. Uncertainties about popula-
assess the probabilities except for Alemi and Agliato tion densities, existing ambient concentrations, and
(1989), Betaque and Gorry (1971), and Mehrez and dose/response relationships are considered along with
Gafni (1987). Roughly two-thirds of the studies use a the values of U.S. Environmental Protection Agency
single evaluation measure with a decision tree. Note that
decision makers.
Clarke (1987), Fryback and Keeney (1983), Mehrez and Ulvila and Snider (1980) model the international nego-
Gafni (1987), and Pliskin, Shepard and Weinstein (1980)tiating situation faced by the U.S. in attempting to
propose models for decision making but do not apply negotiate standards for marine pollution. Role playing by
them to actual applications. The remainder of the articles U.S. Coast Guard personnel as well as special interest
involving specific models report actual results from ap- groups is used to determine tradeoffs among multiple
plications. In addition, Sanchez de Rivera (1980) and objectives to use in modeling the positions of various
Venta and Venta (1987) are able to develop simple countries involved in the negotiations. These are used to
decision rules as a result of their analyses of jaundiceidentify compromise solutions in the standards setting
and deep-vein thrombosis, respectively, while Betaque negotiations that might be acceptable to various coun-
and Gorry (1971) report the development and application tries. Van Steen (1987) develops a procedure to analyze
of a computerized aid for diagnosis. the desirability of allowing transportation of certain
classes of hazardous materials through tunnels in The
Public Policy
Netherlands. The paper argues that decision analysis
The public policy articles can be separated into those offers the appropriate perspective for dealing with ac-
dealing with standard setting and those dealing with ceptable risk problems, and includes an approach for
other miscellaneous issues. providing insight into value judgments required to arrive
Standard Setting. Anandalingam (1987) evaluates acid at a decision.
rain regulation. An objectives hierarchy is developed
including economic, environmental, socioeconomic, and Miscellaneous. Anandalingam and Olsson (1989) apply
other objectives. The intent of this framework is to a three-step methodology to a decision concerning the
structure negotiations between states to identify compro- future fresh water supply for a municipality. Two
mise regulations for acid rain. Anandalingam (1989) screening steps are used to reduce the number of alterna-
extends this work by using multiattribute utility functions tives that need to be considered in the final step which

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
214 / CORNER AND KIRKWOOD

uses a multiattribute value function. Bodily (1978) incor- uncertain degree of the voluntary use of the new zip code
porates the preferences of three interest groups (con- system by business. Ulvila and Seaver (1982) analyze
cerned citizens, administrators, and service personnel) the water supply problems of various Boston suburbs.
into an analysis of the call response time for police They develop an evaluation model including seven crite-
services that takes into account goals for efficiency and ria important to three interested groups of communities
quality of service. Chen et al. (1979) develop a "value affected by a proposed project. Tradeoffs among evalua-
oriented social decision analysis" method to integrate the tion criteria are elicited for each group.
preferences of interested parties into the political deci- Wenstop and Carlsen (1988) focus on a development
sion making process considering the installation of a strategy for 542 proposed Norwegian hydroelectric pro-
solid waste shredding facility. The objective of this jects. A study by the Norwegian Ministry of Environ-
analysis is to make the preferences of the affected parties ment had ranked the proposed projects, where high
clearer to the person making the decision. rankings were achieved by those projects exhibiting low
Cohan et al. (1984) cite three applications of decision cost and minimal conflict. Wenstop and Carlsen deemed
analysis to deciding on prescribed forest burns in the that study to be irrational, thus prompting their multiat-
Southwest United States. Three types of decisions are tribute utility analysis. The article and an accompanying
considered: treatment selection, planning, and execution. letter from officials of the Ministry of Environment
A three component model is used, including decision illustrate the complexities of decision analysis in a
trees to analyze sequential decisions and uncertainties, a political setting.
component reflecting the interaction between humans
General
and the forest, and a value model for tradeoffs among
conflicting objectives. Edwards and von Winterfeldt Nine articles do not fit into the four categories discussed
(1987) present an iterative and interactive procedure above. Dalkey (1981) uses decision analysis in an at-
using multiattribute utility functions for formally includ-tempt to resolve Hamlet's famous decision problem "to
ing the values of various stakeholder groups in public be or not to be." Dyer and Miles (1976) consider
risk debates. The approach is applied to three separate trajectory decisions for the two 1977 Mariner
public policy issues: the future development of Jupiter/Saturn space missions. Collective choice rules
Germany's energy system, offshore oil leasing policies were applied to this large-group decision problem to
for southern California, and the siting of water control select a pair of trajectories to meet science investigation
dams in central Arizona. objectives while keeping within mission constraints.
Manne, Richels and Weyant (1979) review the appli- Farmer (1985) compares several multiattribute utility
cation of decision analysis to government research and approaches for modeling external auditor perceptions of
development policy decisions concerning synthetic fuels the reliability of a firm's internal control systems. Janssen
and breeder reactors. Peck (1980) argues that decision and Daniel (1984) analyze the American football deci-
analysis can be an effective communication tool in con- sion to either run/pass or kick for conversion in the
sidering the effects of both a nuclear power generation closing moments of a close game.
moratorium and funding of advanced nuclear reactor Longbottom and Wade (1973) interviewed personnel
concepts. This illustrative article combines such analysis from seventeen selected United Kingdom firms in the
with other previously developed energy system models. summer of 1971 to investigate the use of decision analy-
Sarin (1983) conducts a decision analysis to formulate sis. They determined that deterministic sensitivity
and evaluate alternative policies to address seismic safety analysis was the primary tool used at that time. Decision
problems faced by the city of Los Angeles with regard to trees, judgmental probability estimates, and risk analysis
its old masonry buildings. Tradeoffs between upgrade methods were also generally used. Capital investment
costs and safety (reduced injuries and deaths in the event was the most common application area, followed by
of an earthquake) are examined from the viewpoint of marketing. Luna and Reid (1986) present a decision tree
building owners, renters, policy makers and planners, analysis approach for individuals attempting to select
and the public-at-large. from available mortgage instruments. Heian and Gale
Ulvila (1987, 1988) found it impractical to assess a (1988) extend Luna and Reid's mortgage selection model
utility function for the U.S. Government, but the results to include the borrower's time preferences and attitude
of his analysis of the new ZIP + 4 automation technol- toward risk.
ogy turned out to be insensitive to uncertainties, making Pearman (1987) provides a recent comparison of U.K.
determination of a utility function unnecessary. Used as and U.S. decision analysis applications. Through inter-
the basis for a report to Congress, the model addresses views and an appraisal of the literature, he found that the
technological and economic uncertainties, as well as the U.K. lags the U.S. in applications of decision analysis.

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
 Decision Analysis Applications Survey / 215

Table II
Articles Addressing Methodological and Implementation Issues

Problem Structuring/Formulation
Anandalingam (1987), Anandalingam (1989), Anandalingam and Olsson (1989), Belton (1985), Betaque and Gorry (1971), Bodily (1977),
Dyer and Lorber (1982), Dyer and Lund (1982), Dyer and Miles (1976), Edwards and von Winterfeldt (1987), Fryback and Keeney (1983),
Keeney (1979), Keeney and Sicherman (1983), Keeney and Smith (1982), Kirkwood (1982), Merkhofer and Keeney (1987), North and Stengel
(1982), Ozernoy, Smith and Sicherman (1981), Peerenboom, Buehring and Joseph (1989), Smallwood and Morris (1980), Ulvila and Snider
(1980), van Steen (1987), Wenstop and Carlsen (1988), Wooler (1982).

Decision Trees

Alemi and Agliato (1989), Bell (1984), Betaque and Gorry (1971), Chen et al. (1979), Clarke (1987), Cohan et al. (1984), Crawford,
Huntzinger and Kirkwood (1978), Digman (1980), Hosseini (1986), Janssen and Daniel (1984), Jensen, Tome and Darby (1989), Judd and
Weissenberger (1982), Keeney (1987), Keeney, Lathrop and Sicherman (1986), Keeney and Ozernoy (1982), Kirkwood (1982), Luna and
Reid (1986), Madden, Hyrnick and Hodde (1983), Manne, Richels and Weyant (1979), North and Stengel (1982), Peck (1980), Phillips
(1982), Pliskin, Ronen and Feldman (1987), Ronen and Pliskin (1981), Ronen, Pliskin and Feldman (1984), Sanchez de Rivera (1980), Sarin
(1983), Shachtman (1980), Smallwood and Morris (1980), Ulvila (1987), Ulvila, Brown and Packard (1977), Venta and Venta (1987), von
Winterfeldt (1982), Wells (1982), Winter (1985), Zalkind and Shachtman (1983).

Probability Assessment

Betaque and Gorry (1971), Crawford, Huntzinger and Kirkwood (1978), Dalkey (1981), Hosseini (1986), Jensen, Tome and Darby (1989),
Keefer and Kirkwood (1978), Keeney, Lathrop and Sicherman (1986), Keeney and Sicherman (1983), Longbottom (1973), Madden, Hyrnick
and Hodde (1983), North and Stengel (1982), Ronen and Pliskin (1981), Sanchez de Rivera (1980), Ulvila (1987), Ulvila, Brown and Packard
(1977).

Utility Assessment

Anandalingam (1989), Anandalingam and Olsson (1989), Betaque and Gorry (1971), Bodily (1978), Brooks and Kirkwood (1988), Crawford,
Dyer and Lorber (1982), Dyer and Lund (1982), Dyer and Miles (1976), Edwards (1977), Farmer (1987), Fryback and Keeney (1983),
Golabi, Kirkwood and Sicherman (1981), Huntzinger and Kirkwood (1978), Keefer and Kirkwood (1978), Keeney (1979), Keeney, Lathrop
and Sicherman (1986), Keeney and Sicherman (1983), Keeney and Smith (1982), Kirkwood (1982), Merkhofer and Keeney (1987), Pliskin,
Ronen and Feldman (1987), Pliskin, Shepard and Weinstein (1980), Ronen and Pliskin (1981), Ronen, Pliskin and Feldman (1984), Sanchez
de Rivera (1980), Torrance, Boyle and Horwood (1982), van Steen (1987), Zalkind and Shachtman (1983).

Communication/Facilitation
Alemi and Agliato (1989), Belton (1985), Chen et al. (1979), Dyer and Lund (1982), Dyer and Miles (1976), Edwards and von Winterfeldt
(1987), Gregory and Lichtenstein (1987), Keefer and Kirkwood (1978), Keeney (1987), Lock (1982), Madden, Hyrnick and Hodde (1983),
Merkhofer and Keeney (1987), Peck (1980), Peerenboom, Buehring and Joseph (1989), Phillips (1982), Pollock and Chen (1986), Smallwood
and Morris (1980), Ulvila (1988), Ulvila, Brown and Packard (1977), Ulvila and Snider (1980), Wells (1982), Wenstop and Carlsen (1988),
Winter (1985).

Group Decision Making

Bodily (1978), Dyer and Miles (1976), Edwards (1977), Edwards and von Winterfeldt (1987), Golabi, Kirkwood and Sicherman (1981),
Keeney, Lathrop and Sicherman (1986), Lathrop and Watson (1982), Lincoln and Rubin (1979), Sarin (1983), Sarin, Sicherman and Nair
(1978), Torrance, Boyle and Horwood (1982), Ulvila and Snider (1980), von Winterfeldt (1982).

As reasons for this, he cites: 1) the natural time lag in Finally, Pollock and Chen (1986) found difficulties in
the dissemination of the method because it was primarily attempting to apply decision analysis to the "black stink"
developed in the U.S., 2) more extensive decision analy- water pollution problem in Shanghai, China. Generation
sis education in the U.S. due to the large number of of action alternatives and accounting for uncertainties
M.B.A. programs, and 3) cultural and organizational was almost impossible because of cultural pressures to
differences which make the U.S. more receptive to the avoid confrontation by not deviating from previously
method. stated or assumed goals.

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
216 / CORNER AND KIRKWOOD

3. METHODOLOGICAL AND IMPLEMENTATION REFERENCES

ISSUES
ALEMI, F., AND J. AGLIATO. 1989. Restricting Patients'
Many of the articles surveyed provide significant detail Choices of Physicians: A Decision Analytic Evaluation

about decision analysis methodological or implementa- of Costs. Interfaces 19 (2), 20-28.

tion issues. Specifically, several articles address each of ALLETT, E. J. 1986. Environmental Impact Assessment and
Decision Analysis. J. Opnl. Res. Soc. 37, 901-910.
the following issues: problem structuring/formulation,
AMEMIYA, T. 1981. Qualitative Response Models: A
decision trees, probability and utility assessment, com-
Survey. J. Econ. Lit. 19, 1483-1536.
munication/facilitation, and group decision making.
ANANDALINGAM, G. 1987. A Multiple Criteria Decision
Table II shows articles that address each issue.
Analytic Approach for Evaluating Acid Rain Policy
Articles are included in the problem structuring/
Choices. Eur. J. Opnl. Res. 29, 336-352.
formulation category if they provide an objectives hierar-
ANANDALINGAM, G. 1989. A Multiagent Multiattribute Ap-
chy and/or explain the structuring process in significant proach for Conflict Resolution in Acid Rain Impact
detail. Articles are included in the decision tree category Mitigation. IEEE Trans. Syst. Man Cybernet. 19,
if they pictorially present a decision tree and discuss its 1142-1153.
use. Articles are included in the probability assessment ANANDALINGAM, G., AND C. E. OLSSON. 1989. A Multi-
category if probabilistic dependence/independence ques- Stage Multi-Attribute Decision Model for Project Se-
tions are discussed and/or there is substantial discussion lection. Eur. J. Opnl. Res. 43, 271-283.
of judgmentally elicited distributions. Likewise, articles BELL, D. E. 1984. Bidding for the S.S. Kuniang. Inter-

in the utility assessment category discuss lottery ques- faces 14 (2), 17-23.

tions to elicit utility functions, pictorially present the BELL, D. E., R. L. KEENEY AND H. RAIFFA (EDS.). 1977.

resulting utility functions, and/or explicitly describe Conflicting Objectives in Decisions. John Wiley,
Chichester, U.K.
assessment of tradeoffs among evaluation measures.
BELTON, V. 1985. The Use of a Simple Multiple-Criteria
Those articles included in the communication/facilita-
Model to Assist in Selection From a Shortlist. J.
tion category consider such issues as the success or
Opnl. Res. Soc. 36, 265-274.
failure of the analysis, reflections on the analysis by
BEN-AKIVA, M., AND S. R. LERMAN. 1985. Discrete Choice
involved parties, or the role of the analyst. Finally,
Analysis: Theory and Applications to Travel De-
articles included in the group decision making category mand. MIT Press, Cambridge, Mass.
deal with aggregation of several individual's preferences BETAQUE, N. E., AND G. A. GORRY, 1971. Automating
into a single function. Judgmental Decision Making for a Serious Medical
Problem. Mgmt. Sci. 17, B-421-B-434.
BODILY, S. E. 1977. A Multiattribute Decision Analysis for
4. CONCLUDING COMMENTS the Level of Frozen Blood Utilization. IEEE Trans.
Syst. Man Cybernet. SMC-7, 683-694.
This survey provides a guide to decision analysis appli-
BODILY, S. 1978. Police Sector Design Incorporating Pref-
cations appearing in major English language operations
erence of Interest Groups for Equality and Efficiency.
research journals and other closely related journals from
Mgmt. Sci. 24, 1301-1313.
1970 through 1989. These applications cover a wide
BROOKS, D. G., AND C. W. KIRKWOOD. 1988. Decision
range of decision problems in both the public and private
Analysis to Select a Microcomputer Networking Strat-
sectors. It is worth reiterating the comment made at the egy: A Procedure and a Case Study. J. Opnl. Res.
beginning of the article that many of the applications Soc. 39, 23-32.
address strategic or policy decisions. Thus, they provide CELONA, J., AND P. McNAMEE. 1988. SUPERTREE
counterexamples to the often-stated criticism that opera- User's Guide, 2nd ed. Strategic Decisions Group,
tions research addresses lower level operational prob- Menlo Park, Calif.
lems successfully, but not the strategic issues that are the CHEN, K., J. C. MATHES, K. JARBOE AND J. WOLFE. 1979.
central concern of top management in business and Value Oriented Social Decision Analysis: Enhancing
government. Mutual Understanding to Resolve Public Policy Issues.
IEEE Trans. Syst. Man Cybernet. SMC-9, 567-580.
CHEUNG, D. C., AND C. W. KIRKWOOD. 1989. EXPRES-
ACKNOWLEDGMENT SION TREE: A Decision Tree Analyzer With Vari-
ables and Expressions. Technical Report DIS 88/89-9,
We wish to thank Thomas E. Sandman for his assistance Department of Decision and Information Systems,
in preparing material for this article. Daniel G. Brooks Arizona State University, Tempe.
and Donald L. Keefer made helpful suggestions concern- CLARKE, J. 1987. The Application of Decision Analysis to
ing the presentation. Clinical Medicine. Interfaces 17 (2), 27-34.

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
 Decision Analysis Applications Survey / 217

COHAN, D., S. M. HAAS, D. L. RADLOFF AND R. F. HIGGINS, J. C. 1982. Decision-Making at Board Level
YANCIK. 1984. Using Fire in Forest Management: Using Decision Analysis: Two Case Studies. J. Opnl.
Decision Making Under Uncertainty. Interfaces 14 Res. Soc. 33, 319-326.
(5), 8-19. HOBBS, B. F. 1980. A Comparison of Weighting Methods
Committee on The Next Decade in Operations Research in Power Plant Siting. Dec. Sci, 11, 725-737.
(CONDOR). 1988. Operations Research: The Next HOLLOWAY, C. A. 1979. Decision Making Under Un-
Decade. Opns. Res. 36, 619-637. certainty: Models and Choices. Prentice-Hall,
CRAWFORD, D. M., B. C. HUNTZINGER AND C. W. Englewood Cliffs, N.J.
KIRKWOOD. 1978. Multiobjective Decision Analysis for HOSSEINI, J. 1986. Decision Analysis and Its Application in
Transmission Conductor Selection. Mgmt. Sci. 24, the Choice Between Two Wildcat Oil Ventures. Inter-
1700-1709. faces 16 (2), 75-85.
DALKEY, N. C. 1981. A Case Study of a Decision Analysis: HOWARD, D. A. 1988. Decision Analysis: Practice and
Hamlet's Soliloquy. Interfaces 11 (5), 45-49. Promise. Mgmt. Sci. 34, 679-695.
DAWES, R. M. 1988. Rational Choice in an Uncertain HowAR.D, R. A., AND J. E. MATHESON. 1984a. Influence
World. Harcourt Brace Jovanovich, San Diego. Diagrams. In Readings on the Principles and Appli-
DIGMAN, L. A. 1980. A Decision Analysis of the Airline cations of Decision Analysis, Vol. II: Professional
Coupon Strategy. Interfaces 10 (2), 97-101. Collection, R. A. Howard and J. E. Matheson (eds.).
DYER, J. S., AND H. W. LORBER. 1982. The Multiattribute Strategic Decisions Group, Menlo Park, Calif.
Evaluation of Program-Planning Contractors. Omega HowARD, R. A., AND J. E. MATHESON (EDS.). 1984b.
10, 673-678. Readings on the Principles and Applications of De-
DYER, J. S., AND R. N. LUND. 1982. Tinker Toys and cision Analysis. Strategic Decisions Group, Menlo
Christmas Trees: Opening a New Merchandising Pack- Park, Calif.
age for Amoco Oil Company. Interfaces 12 (6), JANSSEN, C. T. L., AND T. E. DANIEL. 1984. A Decision
38-52. Theory Example in Football. Dec. Sci. 15, 253-259.
DYER, J. S., AND R. F. MILES, JR. 1976. An Actual JENSEN, D. D., A. E. ToME AND W. P. DARBY. 1989.
Application of Collective Choice Theory to the Selec- Applying Decision Analysis to Determine the Effect of
tion of Trajectories for the Mariner Jupiter/Saturn Smoke Detector Laws on Fire Loss in the United
1977 Project. Opns. Res. 24, 220-244. States. Risk Analysis 9, 79-89.
EDWARDS, W. 1977. How to Use Multiattribute Utility JUDD, B. R., AND S. WEISSENBERGER. 1982. A Systematic
Measurement for Social Decision Making. IEEE Approach to Nuclear Safeguards Decision-Making.
Trans. Syst. Man Cybernet. SMC-7, 326-340. Mgmt. Sci. 28, 289-302.
EDWARDS, W., AND D. VON WINTERFELDT. 1987. Public KEEFER, D. L., AND S. E. BODILY. 1983. Three-Point
Values in Risk Debates. Risk Analysis 7, 141-158. Approximations for Continuous Random Variables.
FARMER, T. A. 1987. Testing the Robustness of Multiat- Mgmt. Sci. 29, 595-609.
tribute Utility Theory in an Applied Setting. Dec. Sci. KEEFER, D. L., AND C. W. KIRKWOOD. 1978. A Multiob-
18, 178-193. jective Decision Analysis: Budget Planning for Product
FORST, B. E. 1974. Decision Analysis and Medical Mal- Engineering. J. Opnl. Res. Soc. 29, 435-442.
practice. Opns. Res. 22, 1-12. KEENEY, R. L. 1977. The Art of Assessing Multiattribute
FRENCH, S. 1989. Readings in Decision Analysis. Utility Functions. Orgn. Behavior and Human
Chapman and Hall, London. Perform. 19, 267-310.

FRYBACK, D. G., AND R. L. KEENEY. 1983. Constructing a KEENEY, R. L. 1979. Evaluation of Proposed Storage Sites.
Complex Judgmental Model: An Index of Trauma Opns. Res. 27, 48-64.
Severity. Mgmt. Sci. 29, 869-883. KEENEY, R. L. 1982. Decision Analysis: An Overview.
GOLABI, K., C. W. KIRKWOOD AND A. SICHERMAN. 1981. Opns. Res. 30, 803-838.
Selecting a Portfolio of Solar Energy Projects Using KEENEY, R. L. 1987. An Analysis of the Portfolio of Sites
Multiattribute Preference Theory. Mgmt. Sci. 27, to Characterize for Selecting a Nuclear Repository.
174- 189. Risk Analysis 7, 195-218.
GREGORY, R., AND S. LICHTENSTEIN. 1987. A Review of KEENEY, R. L. 1988. Building Models of Values. Eur. J.
the High-Level Nuclear Waste Repository Siting Anal- Opnl. Res. 37, 149-157.
ysis. Risk Analysis 7, 219-223. KEENEY, R. L., J. F. LATHROP AND A. SICHERMAN. 1986.
HARVEY, C. M. 1983. Cost-Benefit Study of a Proposed An Analysis of Baltimore Gas and Electric Company's
Daylight Running Lights Safety Programme. J. Opnl. Technology Choice. Opns. Res. 34, 18-39.
Res. Soc. 34, 37-43. KEENEY, R. L., AND V. M. OZERNOY. 1982. An Illustrative
HEIAN, B. C., AND J. R. GALE. 1988. Mortgage Selection Analysis of Ambient Carbon Monoxide Standards. J.
Using a Decision-Tree Approach: An Extension. Opnl. Res. Soc. 33, 365-375.
Interfaces 18 (4), 72-83. KEENEY, R. L., AND H. RAIFFA. 1976. Decisions with

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
218 / CORNER AND KIRKWOOD

Multiple Objectives: Preferences and Value Trade- MERKHOFER, M. W. 1987. Quantifying Judgmental Uncer-
offs. John Wiley, New York. tainty: Methodology, Experiences, and Insights. IEEE
KEENEY, R. L., AND A. SICHERMAN. 1983. Illustrative Trans. Syst. Man Cybernet. SMC-17, 741-752.
Comparison of One Utility's Coal and Nuclear Choices. MERKHOFER, M. W., AND R. L. KEENEY. 1987. A Multiat-
Opns. Res. 31, 50-83. tribute Utility Analysis of Alternative Sites for the
KEENEY, R. L., AND G. R. SMITH. 1982. Structuring Objec- Disposal of Nuclear Waste. Risk Analysis 7, 173- 194.
tives for Evaluating Possible Nuclear Material Control MILLER, A. C., AND T. R. RICE. 1983. Discrete Approxi-
and Accounting Regulations. IEEE Trans. Syst. Man mations of Probability Distributions. Mgmt. Sci. 29,
Cybernet. SMC-12, 743-750. 352-362.
KELLER, L. R., AND J. L. Ho. 1988. Decision Problem NORTH, D. W., AND D. N. STENGEL. 1982. Decision Anal-
Structuring: Generating Options. IEEE Trans. Syst. ysis of Program Choices in Magnetic Fusion Energy
Man Cybernet. SMC-18, 715-728. Development. Mgmt. Sci. 28, 276-288.
KIRKWOOD, C. W. 1982. A Case History of Nuclear Power OZERNOY, V. M., D. R. SMITH AND A. SICHERMAN. 1981.
Plant Site Selection. J. Opnl. Res. Soc. 33, 353-363. Evaluating Computerized Geographic Information Sys-
KIRKWOOD, C. W. 1990. A Tutorial Summary of tems Using Decision Analysis. Interfaces 11 (5),
Methods for Applied Decision Analysis, Technical 92-99.
Report DIS 89/90-7, Department of Decision and PEARMAN, A. D. 1987. The Application of Decision Analy-
Information Systems, Arizona State University, sis: A US/UK Comparison. J. Opnl. Res. Soc. 9,
Tempe. 775-783.
KIRKWOOD, C. W., AND R. K. SARIN. 1985. Ranking With
PECK, S. C. 1980. Communicating Model Based Informa-
Partial Information: A Method and an Application.
tion for Energy Debates: Two Case Studies. Inter-
Opns. Res. 33, 38-48.
faces 10 (5), 42-48.
KRISCHER, J. P. 1980. An Annotated Bibliography of Deci-
PEERENBOOM, J. P., W. A. BUEHRING AND T. W. JOSEPH.
sion Analytic Applications to Health Care. Opns. Res.
1989. Selecting a Portfolio of Environmental Programs
28, 97-113.
for a Synthetic Fuels Facility. Opns. Res. 37,
LATHROP, J. W., AND S. R. WATSON. 1982. Decision
689-699.
Analysis for the Evaluation of Risk in Nuclear Waste
PHILLIPS, L. D. 1982. Requisite Decision Modeling: A Case
Management. J. Opnl. Res. Soc. 33, 407-418.
Study. J. Opnl. Res. Soc. 33, 303-311.
LINCOLN, D. R., AND E. S. RUBIN. 1979. Cross-Media
PLISKIN, J. S., AND N. PLISKIN. 1980. Decision Analysis in
Environmental Impacts of Coal-Fired Power Plants:
Clinical Practice. Eur. J. Opnl. Res. 4, 153-159.
An Approach Using Multi-Attribute Utility Theory.
PLISKIN, J. S., B. RONEN AND S. FELDMAN. 1987. Choosing
IEEE Trans. Syst. Man Cybernet. SMC-9, 285-289.
an Electronic Cardiac Pacemaker: A Decision Analy-
LOCK, A. R. 1982. A Strategic Business Decision With
sis. Eur. J. Opnl. Res. 32, 333-339.
Multiple Criteria: The Bally Men's Shoe Problem. J.
PLISKIN, J. S., D. S. SHEPARD AND M. C. WEINSTEIN. 1980.
Opnl. Res. Soc. 33, 327-332.
Utility Functions for Life Years and Health Status.
LONGBOTTOM, D. A. 1973. The Application of Decision
Opns. Res. 28, 206-224.
Analysis to a New Product Planning Decision. Opnl.
Res. Quart. 24, 9-17.
POLLOCK, S. M., AND K. CHEN. 1986. Strive to Conquer
the Black Stink: Decision Analysis in the People's
LONGBOTTOM, D., AND G. WADE. 1973. An Investigation
Republic of China. Interfaces 16 (2), 31-37.
into the Application of Decision Analysis in United
Kingdom Companies. Omega 1, 207-215. PRATT, J. W., H. RAIFFA, AND R. SCHLAIFER. 1964. The

LUNA, R. E., AND R. A. REID. 1986. Mortgage Selection Foundations of Decision Under Uncertainty: An Ele-

Using a Decision Tree Approach. Interfaces 16 (3), mentary Exposition. J. Am. Statist. Assoc. 59,

73-81. 353-375.

MADDEN, T. J., M. S. HYRNICK AND J. A. HODDE. 1983. RONEN, B., AND J. S. PLISKIN. 1981. Decision Analysis in
Decision Analysis Used to Evaluate Air Quality Con- Microelectronic Reliability: Optimal Design and Pack-

trol Equipment for Ohio Edison Company. Interfaces aging of a Diode Array. Opns. Res. 29, 229-242.
13 (1), 66-75. RONEN, B., J. S. PLISKIN AND S. FELDMAN. 1984. Balanc-

MANNE, A. S., R. G. RiCHELS AND J. P. WEYANT. 1979. ing the Failure Modes in the Electronic Circuit of a
Energy Policy Modeling: A Survey. Opns. Res. 27, Cardiac Pacemaker: A Decision Analysis. J. Opnl.
1-36. Res. Soc. 35, 379-387.
McNAMEE, P., AND J. CELONA. 1987. Decision Analysis SAMSON, D. 1988. Managerial Decision Analysis. Irwin,
for the Professional- With SUPER TREE. Scientific Homewood, Illinois.
Press, Redwood City, Calif. SANCHEZ DE RIVERA, D. P. 1980. A Decision Analysis
MEHREZ, A., AND A. GAFNI. 1987. The Optimal Treatment Model for a Serious Medical Problem. Mgmt. Sci.
Strategy-A Patient's Perspective. Mgmt. Sci. 33, 26, 707-718.
1602-1612. SARIN, R. K. 1983. A Social Decision Analysis of the

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms
 Decision Analysis Applications Survey / 219

Earthquake Safety Problem: The Case of Existing Los of Multiattribute Value Theory. Opns. Res. 28,
Angeles Buildings. Risk Analysis 3, 35-50. 81-96.
SARIN, R. K., A. SICHERMAN AND K. NAIR. 1978. VAN STEEN, J. F. J. 1987. A Methodology for Aiding
Evaluating Proposals Using Decision Analysis. IEEE Hazardous Materials Transportation Decisions. Eur.
Trans. Syst. Man Cybernet. SMC-8, 128-131. J. Opnl. Res. 32, 231-244.
SHACHTMAN, R. H. 1980. Decision Analysis Assessment of VENTA, E. R., AND L. A. VENTA. 1987. The Diagnosis of
a National Medical Study. Opns. Res. 28, 44-59. Deep-Vein Thrombosis: An Application of Decision
SMALLWOOD, R. D., AND P. A. MORRIS, 1980. A Task Analysis. J. Opnl. Res. Soc. 38, 615-624.
Force Decision Analysis. Opns. Res. 28, 60-80. VON WINTERFELDT, D. 1982. Setting Standards for Offshore
SPETZLER, C. S., AND C. A. S. STAEL VON HOLSTEIN. 1975. Oil Discharges: A Regulatory Decision Analysis.
Probability Encoding in Decision Analysis. Mgmt. Opns. Res. 30, 867-886.
Sci. 22, 340-358. VON WINTERFELDT, D., AND W. EDWARDS. 1986. Decision
Texas Instruments, Inc. 1985. Arborist Decision Tree Analysis and Behavioral Research. Cambridge Uni-
Software User's Guide. Texas Instruments, Inc., versity Press, Cambridge, U.K.
Austin. WAGNER, H. M., M. H. ROTHKOPF, C. J. THOMAS AND H.
TORRANCE, G. W., M. H. BOYLE AND S. P. HARWOOD. J. MISER. 1989. The Next Decade in Operations Re-
1982. Application of Multi-Attribute Utility Theory to search: Comments on the CONDOR Report. Opns.
Measure Social Preferences for Health States. Opns. Res. 37, 664-672.
Res. 30, 1043-1069. WELLS, G. E. 1982. The Use of Decision Analysis in
ULVILA, J. W. 1987. Postal Automation (Zip + 4) Technol- Imperial Group. J. Opnl. Res. Soc. 33, 313-318.
ogy: A Decision Analysis. Interfaces 17 (2), 1-12. WENSTOP, F. E., AND A. J. CARLSEN. 1988. Ranking
ULVILA, J. W. 1988. 20/30 Hindsight: The Automatic Hydroelectric Power Projects With Multicriteria Deci-
Zipper. Interfaces 18 (1), 74-77. sion Analysis. Interfaces 18 (4), 36-48.
ULVILA, J. W., AND R. V. BROWN. 1982. Decision Analysis WINTER, F. W. 1985. An Application of Computerized
Comes of Age. Harvard Bus. Rev. 60, 130-141. Decision Tree Models in Management-Union Bargain-
ULVILA, J. W., R. V. BROWN AND K. S. PACKARD. 1977. A ing. Interfaces 15 (2), 74-80.
Case in On-line Decision Analysis for Product Plan- WOOLER, S. 1982. A Decision Aid for Structuring and
ning. Dec. Sci. 8, 598-615. Evaluating Career Choice Options. J. Opnl. Res.
ULIVA, J. W., AND D. A. SEAVER. 1982. Decision Analysis Soc. 33, 343-351.
for Water Resources Planning. Omega 10, 185-194. ZALKIND, D. L., AND R. H. SHACHTMAN. 1983. A Method
ULVILA, J. W., AND W. D. SNIDER. 1980. Negotiation of to Determine a Non-Economic Personal Value of Life.
International Oil Tanker Standards: An Application J. Opnl. Res. Soc. 34, 145-153.

This content downloaded from

103.10.28.179 on Tue, 09 May 2023 14:20:09 +00:00
All use subject to https://about.jstor.org/terms