Welcome to Practical Stats

For several decades we provided statistical training and consulting for the environmental sciences and natural resources. Now that our lead scientist has retired, we are making available here a number of resources for you to freely view so that you might "Make Sense of Your Data" (our motto).

Ending 12/31/2026 -- Our Nondetects And Data Analysis training course Note that this course will be discontinued on Dec 31, 2026. Most people work on one lesson per week, taking 13 weeks to complete the course. If you start this course on or before Oct 1, 2026 you should be able to complete it. Until the end of 2026 the NADA course continues to be available at NO COST on this website. It successfully trained students over many years in statistical methods for data with nondetects. All materials are here, but no email support is provided. Questions by email will likely never be answered. But hey, its now free!

Videos

-- Recordings of webinars given by Dennis Helsel over the years. Our "𝝁Tube" for environmental statistics.
Available in the Videos menu are:
Videos for data with nondetects (censored data)
Videos for general environmental statistics.

Textbooks

1. The 2020 version of the classic textbook Statistical Methods in Water Resources is available as
* a free download. See the link on our Books page
* an inexpensive hardcopy sold by the US Geological Survey, at https://store.usgs.gov/product/533012

2. Statistics for Censored Environmental Data (2nd edition) was published by Wiley in 2012 and is available for purchase as an ebook or print.

Software

The NADA2 package for R provides advanced data analysis methods for censored data. It is now freely available at https://cran.r-project.org/package=NADA2 .
Now available as NADA2 v. 2.0 or higher on the CRAN site.

NADA2 goes well beyond estimating summary statistics in providing methods for data analysis of data with one or multiple detection limits. Without substituting any fraction of the detection limit(s), using NADA2 you can:
* Compute confidence, prediction and tolerance intervals
* Test for exceedances of a standard
* Perform matched pair (before/after, upstream/downstream, etc.) tests
* Test for group differences -- parametric, nonparametric and permutation tests
* Compute correlation coefficients and regression
* Draw QQ and partial plots to guide whether to transform x and y variables in regression
* Build multiple regression models and find models with the least error
* Perform parametric and nonparametric trend analyses. For example, the centrendsea function performs a nonparametric Seasonal Kendall trend analysis for censored data after first
adjusting for the effect of a covariate (a variable that is not time)
* Cluster groups of censored multivariate data
* Draw NMDS plots of censored multivariate data
* Test for group differences and trends of censored multivariate data
… .. and much more.

Newsletters

Newsletters containing bits of information on environmental statistics were published periodically from 2003 - 2021. They are available for download from our Newsletter Archive.

Privacy Policy

This site stores no cookies on your device. It does not track you in any way.
There are no advertisements.
Graphs only display the graphic, with no hidden functions.

Arrow to go to the top of the page