What’s the difference between accuracy and precision?

by | Sep 1, 2021 | Reports

Workers at SVL prepare samples in lab

Have you ever wondered what the difference between accuracy and precision is? The classic example is this: I shoot five arrows from a hundred yards and they land either in or just outside of the bullseye.  The consensus from the gallery will be that I am a very accurate shooter.  The next shooter comes up and shoots five arrows and they land low on the target about two feet from the bulls eye but within inches of each other.  These arrows were not very accurate but the shooter showed amazing precision because all the arrows were grouped together.  

What can you take away from this classic explanation?  You may receive an analytical report where all the reproducible QC is spot on but the results don’t reflect the concentration of the analyte in your sample.

What’s the definition of accuracy?

The degree of agreement between an observed value and an accepted reference value.  Accuracy includes a combination of random error (precision) and systemic error (bias) components that are due to sampling and analytical operations; a data quality indicator.

What’s the definition of precision?

The degree to which a set of observations or measurements of the same property, obtained under similar conditions, conform to themselves; a data quality indicator.  Precision is usually expressed as standard deviation, variance or range, in either absolute or relative terms.

What’s an example of where the difference is important?

Nitrate has been deemed by the EPA to be immediately dangerous to life and health at concentration levels above 10 ppm.  You receive an analytical report that indicates that your sample contained nitrate at 9 ppm and also contained 9 ppm in the duplicate of your sample the lab analyzed.  How do you know this is accurate based solely on the analytical report – you don’t.  All the report proved was the analytical precision the lab reported was well done.  The report may not have been accurate, the actual concentration was 11 ppm when one looked at the error associated with the analytical run (which the lab did not provide you with).  So you really had a sample that contained nitrate at a level that the EPA states is dangerous.

How does that difference come up in SVL Analytical’s work?

When you look at your analytical report you should look at the QC section of the report and see how well the Laboratory Control Sample (LCS) was recovered.  For example you see that the LCS was recovered at 95%, you could then apply reduction to your sample.  You should remember though that the LCS is a known concentration put into pristine water, where there are no interferences with analytical recovery.  Your sample may contain matrix interferences that further adjust what is really in your sample.  So to be safe any analytical result close to levels the EPA sets should take into account accuracy over precision.

Join us for more information

about the environmental laboratory industry, plus events we’re attending and hosting.

Biotic Ligand Model

The Biotic Ligand Model (BLM) takes into account many environmental conditions in order to determine if a metal exists in a form or concentration that is toxic to aquatic life.  Toxicity is usually based upon the dosage of whatever it is that we are looking at.  At...

SVL Analytical celebrates 50 years of science, family and mining

From geology exploration to environmental compliance, SVL attributes its longevity to hard work and an adaptive mentality. The laboratory may be an unrecognized partner—that is until something goes wrong like a spill, a breach, a new compliance rule. Our data users...

TCLP or SPLP on a Solid Sample

Why is this TCLP/SPLP on a solid sample a common question? The main reason is that people don’t understand the difference between characterization and hazardous determination.  The EPA developed the TCLP Program to determine when certain materials have to go to a...

What does Cation/Anion Balance Mean?

Introduction to Cation/Anion Balance (CA Balance) Water that is potable (drinkable) should be electrically neutral.  This is determined by taking into account results from the following analytical testing: pH, conductivity, total dissolved solids, and other analytical...

Understanding Control Charts

What are control charts? Control charts are a visual depiction of ongoing data points that various statistics are applied to.  The EPA used to require a laboratory to adjust its spike acceptance recovery percentage based upon how spikes were recovered over time...

How Low Can You Go – Well No Lower Than The MDL

There is always confusion on how low a lab can go when it comes to what their instruments can detect.  The EPA and TNI came together and changed how Method Detection Limits (MDL) are established for an instrument or group of instruments. They wanted to account for the...

Significant Figures Use by Laboratories

What determines the amount of numbers you see on your report? Each test that we do has a level of sensitivity associated with it.  Tests that you can do by visual inspection have the lowest sensitivity and are usually reported in whole numbers in a range of say 1 to...

New Drinking Water COC for Coliform Bacteria Available

We’ve updated our Drinking Water COC/Submittal Form for Coliform Bacteria. You can download it off of our  Resources page.    

The Argument on “Total” Metals

Why is there still an argument about which digestion method should be reported as “total” metals? When you’re looking at an old QAPP, permit or analytical report, you might have noticed that there were different listings for how the metals were to be identified and...

Exciting changes: a new location

After 16 years in our current location, we a relocating our Coeur d'Alene laboratory service to our Kellogg location and establishing a new sample drop-off location in Hayden. Here are some important dates to keep in mind: March 18, 2021: The last day to drop samples...