ENVIRONMENTAL
   STATISTICS
    TRAINING

  COURSES FOR THE
  NON-STATISTICIAN

    TEACHING THE
 CONCEPTS, NOT THE
     FORMULAS
  US Environmental Protection Agency
  Office of Environmental Information
               OfFICEOf
               ENVIRONMENTAL
               INFORMATION

-------
OBJECTIVES

Environmental Statistics Training has been
designed as a non-technical series of courses
for EPA employees and their state/local
colleagues who either directly (or indirectly
through contracts) face issues of data
analysis, data quality, and interpretation of
environmental data.
BACKGROUND

Many of EPA's decisions are based on careful
examination of pertinent data. While many
EPA offices have their own statistician, it is
imperative that non-statisticans who have to
collect, analyze, and interpret data have a
grasp of the basic concepts that are used in
statisical analysis  in order to better
understand and utilize methods for proper
decision making.

As the Agency increases its reliance on data
and information collected from outside
sources, it is essential to have staff capable of
determining the quality, reliability, and utilitity
of information and data.

    •  Can we trust these data?

    •  How much uncertainty is there ?

    •  Do these meet our needs?

    •  How sure are we ?

    •  Can these data hurt us?

    •  Do these data support us?

    •  Did we collect enough data?

    •  Where are the weaknesses?

    •  Are there  any hidden surprises?

    •  What did we do wrong?

-------
OUR  PHILIOSOPHY

In order for EPA to have staff that can feel
comfortable in a data-rich environment, there is
a need to train staff from all backgrounds to
understand how departures from planned
collection of data lead to analytical and
interpretative problems.  Staff need an
understanding of the requirements of a random
sample or one that is representative of a
complex condition. Formal training in statistics
requires theory, mathematics, and use of
numerous formulas.  Environmental Statistics
Training courses are purposely designed to
circumvent these impediments and provide
practical tools and methods for the  non-
statistician.


THREE  LEVELS OF  COURSES

The full list of current courses are listed on
page 5 of this pamphlet. They are divided into
three levels: Basic Statistical Concepts,
Fundamental Statistical Planning, and
Specialized Statistical Practices. The first level
courses are intended for staff having little
mathematical training yet who are interested in
the broad application of statistics. The second
level builds on the first and concentrates on the
planning aspects of data collection  and
generation. The third level looks at useful
methods to determine the reliability, utility, and
usability of data.
PROSPECTIVE AUDIENCE
Some of the staff who have participated in
Environmental Statistics Training include
Project Officers, Principal Investigators,
Managers, Quality Assurance Specialists,
Remedial Program Managers, Chemists,
Biologists, and Environmental Protection
Specialists.

-------
TRAINERS

While the course content is designed for non-
statisticans, the courses are taught by two of EPA's
veteran statisticians, Barry Nussbaum, and John
Warren. These are the guys who have never met a
datum they didn't like. They have spent considerable
time teaching and are known for their humor and
explanatory skills.
HOW TO  TAKE  A  COURSE
The courses are usually offered to specific offices,
Programs, or Regions so that some of the examples
may be tailored to the situation to make the courses
even more applicable.The courses last a half day
each and may be placed in any order to suit the
organization. Project Officers, with permission from
their supervisors, can count each course as  4 hours
training towards their bi-annual certification
requirements.

To set up a course or a series of courses, contact
Barry 202-566-1493,or John 202-564-6876,  and we
will work with you to design a custom set of courses,
arrange schedules, and make travel  arrangements
as might be needed.

-------
HALF-DAY  COURSES


Basic Statistical Concepts

Introduction to DQA: A Reviewer's Guide
How to review a statistical analysis of data

Interpretation of Environmental Statistics:
Turning data into sense and information

Understanding Statistical Tests
How t-tests work and can go wrong


Fundamental Statistical Planning

Systematic Planning and DQOs
How to use statistics in systematic planning

Planning for Sampling
Choosing the appropriate sampling strategy

Translating DQOs to MQOs
Linking objectives to the needs of a QAPP

DQA: Statistical Methods for Practitioners
A survey of statistical techniques

How to Perform a Valid Survey
What you need to know for a good survey
Specialized Statistical Practices

Using Visual Sample Plan Software
The software is free but how do you use it?

Interpretation of Monitoring Data
It's not the same as random data

Dealing with Censored Data
What to do with less-than-detected values

Interpreting Multivariate Data
Handling more than one variable

The Detection of Outliers
How to identify them safely

-------
For more information
Contact:

Barry D. Nussbaum, D.Sc.
202-566-1493
nussbaum.barry@epa.gov

John Warren, Ph.D.
202-564-6876
warren.john@epa.gov
            WWW.EPA.GOV/OEI
                              OfFKEOf
                              ENVIRONMENTAL
                              INFORMATION
 United States
 Environmental Protection
 Agency
Office of        EPA 240F 10001
Environmental    March 2010
Information

-------