analysis, probit analysis is used in assessing ecological risks and engineering, policy, statistical, testing, and legal aspects of the case.
characterizing drinking water exposure to support pesticide regula- Second, compared to the multilevel review process for many anal-
tory decisionmaking. Additionally, linear and nonlinear regression yses, you have about a half second to answer a question on the
models are used to make inference on half-lives and degradation witness stand. Further, it is likely a trick question, posed by the
rates of chemicals in soil, groundwater, and surface water. somewhat antagonistic opposing lawyer. Third, my testimony hap-
The EPA also tries to get there before a problem arises, and pened to occur several weeks before my own doctoral dissertation
part of the pollution prevention activities include such matters as defense. On relative terms, the defense was a piece of cake. So, if
designing sampling protocols to test carpets and carpeting adhesives any graduate students are reading this, here is one way to make the
for volatile organic chemicals as part of the Green Label program. dissertation less painful. However, I do not recommend it. As they
Overarching many of these programmatic activities is the say, “Do not try this at home.”
requirement for quality data and analysis. The EPA has an active
Quality Office. While we know not all data are created equal, we
Getting Out of the Office and Into the Field
spend considerable effort to make sure the data we use are of known
One of the hallmarks of EPA statistics is that the statistician has to
quality and appropriate before their use in applications.
get into the real world to see how the data are accumulated or the
In addition to helping the environment, we use statistical analy-
surveys conducted. I have been to gas stations to view the sampling
sis to help ourselves. Our internal financial operations has a staff
of fuels for vapor pressure and I have seen my own motor vehicle
of analysts working with programs to develop valid, objective, and
tampering survey implemented. Both situations led me to refine
quantifiable measures of performance that cover program activities
the procedures after viewing the actual obstacles to the plan. As
and align them with agency goals. In performing regulatory impact
Yogi Berra said, “You can see a lot just by looking.”
analyses on the EPA’s many regulations, we employ many statisti-
Two other outings are worth noting for their insight. While
cal methods, including revealed preference valuation, uncertainty
viewing a small-business metal plating operation near Detroit,
analysis, precision analysis, switch point statistics, econometrics,
Michigan, it dawned on me that many EPA forms were submitted
and meta-analysis.
by the same employee who did the company’s income tax, work-
men’s compensation, payroll, and cafeteria management. Did this
Enforcement: The Adrenalin Flows
have an effect on data quality? A colleague remembers simply notic-
Among other things, the EPA is an enforcement agency. There are ing flecks of fiber in the cornfields surrounding dumped refractory
times when the statistician must help in preparing a case that might ceramic fibers. The observation led to inducing the manufacturing
ultimately go to trial. For instance, many years ago, I developed company to spend more than $1 million to install scrubbers to
a sampling plan to test in-use vehicles for adherence to emission prevent further releases.
standards. Eventually, the EPA sued Chrysler for excessive carbon
monoxide emissions. I found myself on the stand as the expert sta-
Meta-Analysis and Its Cousin, Meta-Data
tistical witness in the United States v. Chrysler and had to defend Not surprisingly, environmental sampling is expensive. A statisti-
my plan and inference that compelled the EPA to force Chrysler cian cannot always prescribe the perfect experiment and expect to
to recall 208,000 vehicles … and my sample consisted of only 10 implement it. Therefore, meta-analysis has become very important.
cars. The EPA won the case and three aspects are worth mention- Naturally, it is desirable for the EPA to combine several studies
ing. First, this was truly a team effort. All of us had to know the to obtain a more powerful result. To do this frequently requires
an exhaustive examination of the meta-data, the explanation sur-
rounding the original data. Frequently, the meta-data are not com-
plete and the idea of combining studies still looms as a major chal-
lenge for the EPA. While the objective usually is to increase sample
size through meta-analysis, one irony is that there have been times
Nominate a
when the EPA has questioned studies in which data from a study
are subdivided into two studies. Each resultant study now has lower
Member for Fellow
power and one wonders if the study was incorrectly divided so as to
diminish the power of an unwanted result.
The nomination form and guidelines are
Keeping It All Together and What’s in a Name
available at www.amstat.org/fellows.
As may be apparent from this discussion of the many varied prob-
lems and challenges at EPA being addressed by statistical methods,
The deadline for nominating a Fellow is
there are analysts doing this all over the agency. Among other ways
to keep the statistical community informed, we run a monthly EPA
Monday, March 2, 2009.
Statistics Users Group meeting. We are careful at the EPA to call
this a statistics users group, as many, many of the analysts who do
Please be sure that anyone you nominate has been a continuous member of
statistical analysis are not labeled “statisticians.” We want to make
the ASA from March 1, 2006, to February 28, 2009. sure we include all interested parties.
Be sure potential nominees have not already been inducted. To view a
This mixed team approach works well at the EPA. I enjoy work-
searchable list of Fellows, visit www.amstat.org/fellows. ing with all of them, and I certainly thank many of my colleagues
For questions about ASA Fellow nominations, email fellows@amstat.org.
for taking the time and effort to provide me with glimpses into
their work assignments while I prepared this article. n
24 AMSTAT NEWS NOVEMBER 2008
AMSTAT November 08.indd 24 10/24/08 2:28:18 PM
Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52  |  Page 53  |  Page 54  |  Page 55  |  Page 56  |  Page 57  |  Page 58  |  Page 59  |  Page 60  |  Page 61  |  Page 62  |  Page 63  |  Page 64  |  Page 65  |  Page 66  |  Page 67  |  Page 68  |  Page 69  |  Page 70  |  Page 71  |  Page 72  |  Page 73  |  Page 74  |  Page 75  |  Page 76  |  Page 77  |  Page 78  |  Page 79  |  Page 80  |  Page 81  |  Page 82  |  Page 83  |  Page 84