NRI Data Reliability
NRI is a statistically-based survey that
has been designed and implemented using scientific principles to
assess conditions and trends in the state of our natural
resources. The National Resources Inventory is conducted in
cooperation with the Iowa
State University Center for Survey Statistics & Methodology (CSSM).
Margin of Error computations are
automatically generated with most NRI data analyses using the U-CARP
statistical package developed as part of a cooperative agreement with
Iowa State University. A Margin of Error defines the confidence
interval associated with each estimate within the report. All Margin of Error estimates are for the
95% confidence level.
A confidence interval/margin of error has
a convenient interpretation. Specifically, for a 95% confidence
interval, the probability that the interval covers the true mean value
is equal to 0.95. This means that if we draw an infinite number
of samples for a specified population (e.g. cropland acres, erosion
rates, etc.), and if for each sample we compute the confidence
interval, then 95% of those intervals will contain the true mean.
The measure "95%" represents the degree of our confidence
that the interval contains the true population mean.
To make this concept more concrete, consider a
watershed. We want an estimate of the number of acres of cropland in this
watershed. The number of acres of cropland constitutes our population.
From the NRI data, suppose that we estimate the number of cropland acres to be
50,000 and the confidence interval to be (47,000-53,000). This means that
the probability that the actual number of acres of cropland in the watershed de
facto is between 47,000 and 53,000 is 0.95.
The sample design for the 1997 NRI is designed to
provide acceptable margins of error at the 95% confidence level for geographic
areas based on USDA Major Land Resource Area definitions. Analyses based
on smaller geographic areas may produce results with meaningless margins of
Link to the National
NRI website on Statistical Estimation that provides appropriate scientific