Should We Risk It?

by Kammen, Daniel M.; Hassenzahl, David M.

• ISBN13: 9780691074573
• ISBN10: 0691074577
• Edition: Reprint
• Format: Paperback
• Copyright: 2001-03-26
• Publisher: Princeton Univ Pr

Summary

How dangerous is smoking? What are the risks of nuclear power or of climate change? What are the chances of dying on an airplane? More importantly, how do we use this information once we have it? The demand for risk analysts who are able to answer such questions has grown exponentially in recent years. Yet programs to train these analysts have not kept pace. In this book, Daniel Kammen and David Hassenzahl address that problem. They draw together, organize, and seek to unify previously disparate theories and methodologies connected with risk analysis for health, environmental, and technological problems. They also provide a rich variety of case studies and worked problems, meeting the growing need for an up-to-date book suitable for teaching and individual learning. The specific problems addressed in the book include order-of-magnitude estimation, dose-response calculations, exposure assessment, extrapolations and forecasts based on experimental or natural data, modeling and the problems of complexity in models, fault-tree analysis, managing and estimating uncertainty, and social theories of risk and risk communication. The authors cover basic and intermediate statistics, as well as Monte Carlo methods, Bayesian analysis, and various techniques of uncertainty and forecast evaluation. The volume's unique approach will appeal to a wide range of people in environmental science and studies, health care, and engineering, as well as to policy makers confronted by the increasing number of decisions requiring risk and cost/benefit analysis. Should We Risk It? will become a standard text in courses involving risk and decision analysis and in courses of applied statistics with a focus on environmental and technological issues.

Table of Contents

Preface	xv
Acknowledgments	xix
Introduction	3	(28)
Defining Risk	3	(3)
Structure of the Book	6	(2)
Risk Analysis and Public Policy	8	(22)
Getting Started	16	(5)
Data Needs	21	(2)
Using Data	23	(1)
Additional Cases	24	(1)
Additional Curves	24	(1)
Does the Dose Make the Poison?	25	(1)
One in a Million Risks	25	(1)
Surfing and Smoking	26	(1)
Risks of Nuclear Power	26	(4)
References	30	(1)
Basic Models and Risk Problems	31	(52)
Introduction	31	(1)
Basic Modeling	32	(33)
Volatile Organic Emissions from Household Materials: Wallpaper Glue	35	(2)
Indoor Radon Exposure	37	(13)
Revisited	50	(1)
Equilibrium Concentration	50	(1)
Simple PBPK Model--Continuous Dose	51	(5)
Alternative Depictions	56	(1)
PBPK--Finite Dose of Barium	56	(7)
How Much Resolution Is Too Much?	63	(1)
How Much Information Is Needed?	64	(1)
Sensitivity Analysis	64	(1)
Cause and Effect Relationships	65	(4)
Radon and Cancer	65	(4)
Mechanistic Models and Curve Fitting	69	(12)
Conceiving ``Mechanistic'' Models	69	(4)
Using the Wrong Model, Getting the Model Wrong	73	(2)
Empirically Derived Dose Response	75	(2)
Earthquakes versus Traffic Risks	77	(4)
Conclusion	81	(1)
References	82	(1)
Review of Statistics for Risk Analysis	83	(39)
Introduction: Statistics and the Philosophy of Risk Assessment	83	(21)
Average Radon Exposure	85	(2)
Radon Exposures in Different Regions	87	(1)
Working with Data	87	(3)
Mean and Median: Why Worry?	90	(3)
Sample Data Revisited	93	(3)
Hypothesis Testing and Confidence Intervals	96	(6)
Making Decisions	102	(2)
Distributions	104	(17)
Moving Away from Ignorance	104	(7)
Fitting a Model	111	(6)
R2 Versus X2	117	(1)
How Many Bins?	117	(1)
Distributional Models	117	(3)
Fitting the Lognormal Distribution	120	(1)
Dealing with Grouped Data	120	(1)
References	121	(1)
Uncertainty, Monte Carlo Methods, and Bayesian Analysis	122	(31)
Introduction	122	(6)
Measuring the Speed of Light	124	(2)
Energy Forecasts	126	(2)
Forecasting the Impacts of Climate Change	128	(1)
Bayesian Statistics	128	(14)
Interpreting Test Results	132	(2)
Bayesian Experts	134	(1)
Bayesian Analysis of Radon Concentrations	134	(8)
Monte Carlo Analysis	142	(9)
Exposure to Tap Water in the Home	143	(7)
Uncertainty or Incommensurability?	150	(1)
References	151	(2)
Toxicology	153	(46)
Introduction	153	(1)
Critical Assumptions for Modeling Disease	154	(5)
Assumptions Specific to Toxicology	159	(4)
Assumptions Specific to Epidemiology	163	(34)
Test Data and Carcinogenesis: The Kil-EZ Example	166	(8)
Calculating LEDIO	174	(1)
Maximum Tolerated Dose	174	(1)
1,3-Butadiene	174	(1)
Fitting Data to Mechanistic Models: One-Hit, Two-Hit, Two-Stage Problem	175	(14)
The Cost of Better Data	189	(1)
Model-Free Extrapolation	190	(1)
Variation in Cancer Susceptibility	190	(1)
Variation in Sensitivity and Exposure	191	(1)
Additional Data Set	191	(1)
Exact Two-Stage Formulation	191	(3)
Noncarcinogenic Effects: The EPA Approach	194	(2)
Additional Noncancer End Points	196	(1)
Formaldehyde	196	(1)
References	197	(2)
Epidemiology	199	(32)
Introduction	199	(31)
Cigarette Smoking and Cancer	199	(7)
The Heavy Smoker	206	(1)
All Deaths in the United States	206	(1)
Risk in a Time of Cholera	207	(3)
Pooled Data	210	(1)
What Might Be Missing?	211	(1)
Measurement Error	211	(1)
Benzene Revisited: The Pliofilm Cohort Study	211	(3)
Additional Data	214	(1)
One-Hit Model and Epidemiological Data	214	(2)
Additional Data	216	(1)
Catching Cold: Exponential Spread of Disease	216	(2)
Graphical Presentation	218	(1)
Small Groups	218	(1)
The Spread of AIDS: An Empirical Analysis, or, Does the Model Fit the Data?	218	(7)
Double-Blind Study	225	(3)
Side Effects: Test for Safety	228	(1)
Low-Probability Effects	229	(1)
Death by Cheese?	229	(1)
Cancer Clusters--Real or Not?	229	(1)
References	230	(1)
Exposure Assessment	231	(35)
Introduction	231	(33)
Assessment of Exposures and Risks: The ChemLawn Claim	231	(6)
Can Adults and Children Be Treated the Same?	237	(1)
Contaminated Milk	237	(2)
Biomass Fuels and Childhood Disease	239	(2)
Bioaccumulation of Heptachlor in Beef	241	(6)
Sensitive Receptors: Exposure to Children	247	(1)
Exposure via Breast Milk	247	(1)
Tricholoroethylene Exposure at Woburn, Massachusetts	247	(9)
TCE at Woburn: The Big Picture	256	(1)
Probability Distribution for Radon Exposures (or Risks)	257	(1)
PBPK Models and Gender Differences in the Uptake of Benzene	258	(2)
Acme Landfill	260	(4)
References	264	(2)
Technological Risk	266	(38)
Introduction	266	(5)
Lethality of Plutonium	267	(3)
Cassini Spacecraft Reentry Risk	270	(1)
Event Trees and Fault Trees	271	(31)
Simple Pressure Relief System	272	(6)
Additional Fault Tree	278	(1)
Calculations from Fault Trees	278	(1)
Missing Components, Common-Mode Failures, and the Human Element	278	(2)
Additional Fault Trees	280	(1)
Identifying Problems	281	(1)
Anticipating the Unknown	281	(1)
Oleum and the ``Clever-Proofing'' Problem	281	(1)
Coal-Burning Power Plant Emissions	282	(3)
Commercial Nuclear Power Safety: An Empirical Analysis	285	(2)
The Risk from Nuclear Accidents	287	(3)
Long-Term Risks from High-Level Nuclear Waste: A Case of Extreme Uncertainty	290	(5)
Military Fighter Aircraft	295	(4)
Who Decides What's Important?	299	(1)
Risk of Domestic Airplane Flight	299	(2)
Verifast Airlines	301	(1)
Risk of International Airplane Flight	301	(1)
References	302	(2)
Decision Making	304	(49)
Introduction	304	(5)
Comparing Risk Reduction Measures By Dollar Values	304	(4)
Cost Per Life-Year versus One in a Million	308	(1)
Too Much or Too Little Spending?	308	(1)
Mandatory Helmet Laws	309	(1)
Discussion	309	(1)
Is It Worth It?	309	(1)
High-Level Nuclear Waste	309	(1)
Organizing Processes	309	(43)
Event Trees and Decision Analysis	310	(7)
Who Lives There (and Does It Matter)?	317	(1)
Additional Calculations	318	(1)
Additional Uncertainty	318	(1)
Analysis or Abuse? Transmission of Hoof Blister	318	(8)
Dollars and Decisions	326	(1)
Health and Environmental Technology Policy: Superfund Remediation	326	(9)
Discounting or Dodging?	335	(1)
Taking a Viewpoint	336	(1)
Event-Decision Tree	336	(1)
Regulatory Impact Analysis: Where the Rubber Meets the Road	336	(16)
References	352	(1)
Risk Perception and Communication	353	(40)
Introduction	353	(35)
Same Numbers, Different Stories	354	(3)
Opening Dialogue	357	(1)
Explaining Numbers	357	(1)
Framing a Question: Loss or Gain?	358	(2)
A Project to ``Restore'' or ``Improve'' a Wetland	360	(1)
A Little Bit or a Lot? Violent Agreement on the Numbers	360	(1)
Risk Level, Risk Perceptions, and Psychometric Models	361	(2)
Cognitive Maps	363	(1)
Radiation By Any Other Name	364	(1)
Surfers and the Sun Revisited	365	(1)
Ranking the Risks: Are the Experts Right?	366	(3)
Implications of Differing Perspectives	369	(1)
Risk, Trust, and Rationality	369	(1)
Who Is More Concerned?	370	(1)
The Availability Bias	370	(2)
Aggregation	372	(1)
How Intuitive Are Statistics? The Case of Electromagnetic Fields and Cancer	373	(4)
Alternative Interpretations	377	(1)
Use of Point Estimates versus Distributions	377	(4)
Alternative Options	381	(1)
When Is the New Leaf Turned?	381	(1)
What Will They Think It Means?	382	(1)
What To Tell Them?	383	(1)
Alternative Labeling	384	(1)
Situational Differences	384	(1)
Saccharin and Alar: Why the Difference?	384	(3)
When To Spin?	387	(1)
Can or Should ``Zero Risk'' Be a Goal?	387	(1)
Wrapping Up: Putting the Pieces Together	388	(2)
References	390	(3)
Appendix A. Z-Scores	393	(2)
Appendix B. Student's T-test	395	(3)
Appendix C. Chi-Squared Distribution	398	(3)
Index	401