n
STATISTICAL AND TECHNICAL DEFINITIONS RELATED
TO PRODUCT SPECIFICATIONS
Method Detection Limit (MDL):
“The method detection limit (MDL) is
defined as the minimum concentration of a substance that can be measured and
reported with 99% confidence that the analyte concentration is greater than
zero and is determined from analysis of a sample in a given matrix containing
the analyte.”
1
Note that, “As Dr. William Horwitz once stated, ‘In almost all
cases when dealing with a limit of detection or limit of determination, the
primary purpose of determining that limit is to stay away from it.’”
2
1. CFR 40, part 136, appendix B
2. Statistics in Analytical Chemistry: Part 7 – A Review, D. Coleman and
L Vanatta, American Laboratory, Sept 2003, P. 31.
Precision:
Precision is the numerical agreement between two or more
measurements.
3
The precision can be reported as a range for a measurement
(difference between the min and max). It can also be reported as the standard
deviation or the relative standard deviation. It is a measure of how close
together the measurements are, not how close they are to the correct or true
value.
The precision can be very good and the accuracy very bad. This is a useful
measure of the performance of a test method.
3. Skoog, D.A., West, D. M.,
Fundamental of Analytical Chemistry,
2
nd
ed.,
Holt Rinehart and Winston, Inc, 1969, p. 26.
Accuracy:
Accuracy is the nearness of a measurement to the accepted or true
value.
4
The accuracy can be expressed as a range, about the true value, in which
a measurement occurs (i.e. ±0.5 ppm). It can also be expressed as the %
recovery of a know amount of analyte in a determination of the analyte (i.e.
103.5 %).
This is a useful measure and what most customers are interested in when
they want to know about the performance of a test method.
4. Skoog D.A., West D. M.,
Fundamental of Analytical Chemistry,
2
nd
ed.,
Holt Rinehart and Winston, Inc, 1969, p. 26.
Resolution:
Resolution is the smallest discernible difference between any two
measurements that can be made.
5
For meters this is usually how many decimal
places are displayed. (i.e. 0.01). For titrations and various comparators it is the
smallest interval the device is calibrated or marked to (i.e. 1 drop = 10 ppm, 0.2
ppm for a DRT, or ±half a unit difference for an octaslide or color chart). Note
that the resolution many change with concentration or range. In some cases the
resolution may be less than the smallest interval, if it is possible to make a
reading that falls between calibration marks. This is often done with various
comparators.
One caveat is, that resolution has very little relationship to accuracy or
precision. The resolution will always be less than the accuracy or precision but it is not
a statistical measure of how well a method of analysis works. The resolution can be
very very good and the accuracy and precision can be very, very bad! This is not a
useful measure of the performance of a test method.
5. Statistics in Analytical Chemistry: Part 7 – A Review, D. Coleman and
L Vanatta, American Laboratory, Sept 2003, P. 34.
SMART SPECTRO SPECTROPHOTOMETER 05.04
7
