ICP Vendor Comparison

[Spectre66]

I am considering running a series of tests through 3 of the ICP vendors to evaluate service, speed, accuracy and precision. My current plan is to take a large sample of water in a glass jar, run 1 set of tests, refrigerate the glass jar in a dark environment, send another set one month later and then a 3rd set 2 months later. The hope is to see accuracy and precision over the course of a month and a quarter.

Should I be considering another method of doing a side by side comparison over time? Should I consider spiking any variable(s) to see the response?

I appreciate any and all thoughts.

 

[Ranjib]

I am considering running a series of tests through 3 of the ICP vendors to evaluate service, speed, accuracy and precision. My current plan is to take a large sample of water in a glass jar, run 1 set of tests, refrigerate the glass jar in a dark environment, send another set one month later and then a 3rd set 2 months later. The hope is to see accuracy and precision over the course of a month and a quarter.

Should I be considering another method of doing a side by side comparison over time? Should I consider spiking any variable(s) to see the response?

I appreciate any and all thoughts.

there are folks who have already done this (side by side comparison of ati, triton.. and another vendor), the sample on conditions were different (tank water).. what exactly you are looking for?

 

[Spectre66]

I am more interested in the specific details of accuracy and repeatability over time as well as limits and how it relates to commonly accepted target values. I will also be testing the lab in the US out of Colorado.

 

[Randy Holmes-Farley Verified Community Expert]

How would you evaluate accuracy of the results?

 

[Spectre66]

Tbh I am not sure. I had considered spiking a known quantity of a element that I know from previous tests to be at zero for at least two of the vendors.

Open to suggestions.

 

[Spectre66]

That said, I realize a single element or even a few will have some inaccuracy due to my own handling. My curiousity lays more in variability at the same lab over time vs variability with others. In all honesty I am more or less trying to convince myself that one lab is really no different than the other.

 

[Randy Holmes-Farley Verified Community Expert]

Tbh I am not sure. I had considered spiking a known quantity of a element that I know from previous tests to be at zero for at least two of the vendors.

Open to suggestions.

About the only good way is to make or buy a commercial standard. That will be an expensive endeavor either way.

 

[Randy Holmes-Farley Verified Community Expert]

That said, I realize a single element or even a few will have some inaccuracy due to my own handling. My curiousity lays more in variability at the same lab over time vs variability with others. In all honesty I am more or less trying to convince myself that one lab is really no different than the other.

It’s when they differ that you won’t know which is accurate.