Kara
Michelle Morgan, The Development and Evaluation of a Method
for Risk Ranking
Committee: Paul Fischbeck - co-chair (SDS/EPP), Mike DeKay
- co-chair (Heinz/EPP), Baruch Fischhoff (SDS/EPP) and
Granger Morgan (EPP/ECE/Heinz)
Risk
managers are increasingly interested in incorporating
public input into their decisions. This thesis develops
and evaluates a method by which lay groups can provide
such input by ranking risks that have been quantitatively
and qualitatively described by experts on clear, easily
understood "risk summary sheets."
The
information materials were analyzed to determine how narrative
and table-based information affect ranking judgments.
Results show that table-based information increases agreement
among participant rankings as compared to narrative information
alone. Next, a multiattribute model for calculating an
implied risk ranking was evaluated. The model was based
on ten risk attributes: mortality, four morbidity attributes,
and five qualitative attributes (greatest number of deaths
in a single event, time between exposure and effect, quality
of scientific understanding, ability to control, and uncertainty).
Multiple procedures for eliciting the model were compared.
In addition to forming a basis for the recommended model-estimation
procedures, results indicated that participants use the
information from the model to update their holistic rankings.
A
policy capturing approach was used to investigate whether
participants used a consistent strategy for incorporating
risk-attribute information into their rankings. Analysis
showed that participants consistently based their judgments
on the mortality and morbidity attributes, but results
for the remaining attributes were inconclusive. Finally,
a group ranking process that incorporates the information
materials and multiattribute model was evaluated. Results
show consistency between holistic and multiattribute rankings,
agreement between individual and group rankings, and satisfaction
with the process. Results also show that agreement across
groups on the risk rankings is high when this method is
used.
Work
supported by an EPA STAR Graduate Fellowship, NSF grant
SRB-9512023, EPRI grant WO2955-12, CMA and Carnegie Mellon
academic funds. |
|
Hiroshi
Hayakawa, Automobile Risk Perceptions and Insurance-Purchasing
Decisions: A Japan - U.S. Comparison During Deregulation
Committee: Paul S. Fischbeck - co-chair (SDS/EPP), Baruch
Fischhoff - co-chair (SDS/EPP), Granger Morgan (EPP/ECE/Heinz),
and Koichi Okamoto (Toyo Eiwa Women's University)
Japan's
automobile insurance industry was completely deregulated
in July 1998, in accordance with the agreement made by the
Japanese and US governments in 1996. This dissertation was
designed to take advantage of this opportunity and explore
the cross-national differences in risk perceptions and insurance-purchasing
decisions by focusing on the evolution of consumer behavior
in the Japanese automobile insurance market during deregulation.
This
dissertation starts by examining the risk environments faced
by citizens in the two countries, in the domain of traffic
safety. It contrasts accident risks from several points
of view and discusses possible sources of differences in
risk perceptions and insurance-purchasing decisions. In
addition, it provides the context for understanding the
Japanese non-life insurance market, by describing the history
of the regulatory framework and the resulting market structure.
The
thesis also explores cross-national differences in automobile
risk perceptions and insurance-consuming decisions in Japan
and the US before deregulation. It compares survey results
from 42 Japanese subjects in 1997 with those from 74 US
subjects (Austin, 1996). Systematic differences were observed
in reasons given for having automobile insurance, judged
probabilities of accidents, and judged probabilities of
being at fault. Finally, it examines the results from three
surveys the author conducted in Japan in 1997, 1998 and
1999, in order to assess initial impacts of deregulation.
Although major changes in consumer risk perceptions, and
general belief and decision factors about insurance were
not detected, systematic changes in consumer understanding
of insurance deregulation were found.
Work
supported by Mitsui Marine and Fire Insurance Company, Department
of Engineering and Public Policy at Carnegie Mellon University,
and IEA of Japan Company.
|
|
James J. Corbett, Jr., An Assessment of Air Pollution and
Environmental Impacts from International Maritime Transportation
Including Engineering Controls and Policy Alternatives
Committee: Paul Fischbeck - chair (SDS/EPP), Lester Lave
(GSIA/EPP), Granger Morgan (EPP/ECE/Heinz), and Spyros Pandis
(ChemE/EPP)
Air
emissions from ships are shown to be significant at global,
regional and local scales. This work presents a geographically
resolved, global inventory of emissions from commercial
ship engines operating internationally. Global annual NOx
and SOx emissions from ships are estimated to 3.08 Tg (1012
grams) as N, and 4.24 Tg as S, respectively, more than 14
percent of nitrogen emissions from global fuel combustion
sources and more than 16 percent of sulfur emissions from
world petroleum use. Moreover, nearly 70% of ship emissions
occur within 400 km of land regions, and 85% occur north
of the equator. Other air pollutants and carbon dioxide
were estimated in this ship emissions assessment (SEA).
National
inventory of emissions from commercial ships operating in
US waters, indicates that emissions from ships on US inland
rivers equal about 70% of the emissions from ships on all
three US coastlines combined. On the coastlines, oceangoing
vessels account for most of the commercial ship emissions.
Ship
sulfur emissions are included in a global chemical transport
model to quantify the contribution of these ship
Corbett
- continued on pg. 9
_____________________________________________________________________ |
