On-Demand Webinars

ANALYTICAL DETECTS:
Why Subbing One-Half of the Detection Limit is Trouble and
What you Can do Instead

Measurements of trace chemicals in environmental media (water, air, soils, biota) frequently result in values reported only as less than the laboratory reporting limit ("less-thans", "nondetects", and "qualified values"). The most commonly-used method for incorporating nondetects is to substitute one-half the reporting limit and continue as usual. This may obscure patterns and trends that are present, or may create those that are not present in the original data. It is fraught with error.

Two easily-implemented methods are available that avoid the problems with substitution. They can be used to compute summary statistics (mean, median), perform hypothesis tests, and compute regression models. The two methods incorporate data with one detection limit, or can be adapted to work with multiple detection limits.

Webinar Objectives and Highlights:

• Why substituting one-half the detection limit causes problems
• The effects of substitution on data
• The effects of substitution on means, regression, hypothesis tests
• Two easy methods that actually work with nondetect data

What You Will Learn:

• You will be able to summarize (what's a typical value?) data with nondetects correctly.
• You will no longer find trends and differences that are not really there (as substitution can do).
• You will no longer miss trends and differences that ARE really there (as substitution can do).
• You will know of easy methods for nondetects that actually work well.

Whether you want to learn more about statistical methods or managing detection limits, this webinar will provide participants with information that is not available elsewhere on the web.