I send in a sample once every 3 months but my primary purpose is not to measure my tank perimeters, although it does give me data that I do can’t get through my own testing.
My main purpose is to compare my test results to their results
. My goal is to improve my testing methods so I don't have to have ICP testing unless something is going on with my system that I can't pinpoint.
I have just recently been sending my samples to 3 different ICP testing services and comparing my results to all 3 as well as comparing them to each other....There are differences, some of which I would consider "significant"
Getting correct value is important because we make decisions on what to do and often take action…If the measurement is wrong we are taking bad actions!
As there has been much discussion about the ICP testing precision and accuracy....
Examples:
Skeptical Reefkeeping article..
http://packedhead.net/2015/triton-lab-icp-oes-testing-of-a-certified-artificial-saltwater-standard/
Dan’s Reef
There are several other posts related to this topic...This presents a problem which one is the correct value?
There is a great deal to be said about the answer to this question but the basics in my opinion are this:
All testing is made up of two goals,
Accuracy and
Precision….
Accuracy is how close to the “real value” is the result…
Precision is how reproducible are the results. ICP testing is no different. Tests have several components. Read about it here
https://www.inorganicventures.com/accuracy-precision-mean-and-standard-deviation
First is the test device…
Second the test procedure itself…(Weighing, measuring, time, etc.)
Third the sample quality (uniformity, not contaminated etc.)…
Forth the appraiser (person doing the test).
All of these have errors associated with them, both Systematic and Random Errors. In general errors do not compensate, they accumulate all of these components need to be evaluated for error as a whole system. The procedure of Gage Repeatability and Reproducibility, called a Gage R&R is very helpful tool in looking at this accumulative error in the testing results as it relates to
precision of the test based on your target value and your allowable error. (Tolerance values) This is an important number. It tells you how far off the target (nominal value) the measurement can be before you have unacceptable results. With regard to reef water testing this depends essentially on the parameter being tested…for example for Calcium a range of 380-450ppm---Range of 70 a 10% tolerance would be ± 7ppm—20% ± 14ppm…etc. So one might set the tolerance at ± 10ppm from the target value of say 415-420pp and feel confident keeping an adequate amount of Calcium in the reef. On the other hand if the parameter is Iodine the range might be 0- .2ppm---10% is .02ppm---20% .04ppm …And being that Iodine can be toxic at elevated levels your allowable error is much smaller.
I won’t go into the details for the Gage R&R Procedure you can read about it here
https://quality-one.com/grr/ or here
https://www.qualitydigest.com/inside/metrology-article/basics-gauge-rr.html . The math can be quite intimidating but there are a number of free resources that provide Excel Spreadsheets that will do the work for you. …Just Google Gage R&R ANOVA Excel and you will have several options.
I have been using this in my test development projects as well as working to improve my measurements using test kits as well as the Hanna Checkers.
However
Precision alone is not sufficient. The measurement need to have good
Accuracy, but without Precision it is difficult to know if your measurement it is Accurate without taking a lot of measurements and even then there is a limit to the accuracy no matter how many measurements you take. (A lot of Random errors). This is where reference calibration standards are helpful.
Reference Calibration Standards for our hobby (seawater) are not easy to come by…Especially certified ones…and they are expensive!...Here is an example
https://highpuritystandards.com/elements-in-seawater/ (This is the one used in the article by Rich Ross & Dr. Chris Maupin in Skeptical Reefkeeping). There are other ones that you can get but they are not Certified Standards…Like this one
http://cm-distributors.com/product/multi-reference-solution/ The point is you have to have an good target in order to know if your test is giving accurate results. So the precision can be high (good repeatability and reproducibility) but it is precisely wrong!
CONCLUSION
Ok, I did not intend to go down the rabbit hole of testing (but I did), but wanted to give some background for this statement.
It would be very helpful to all of us who are interested in Precise and Accurate results of test parameters for our reefs, or like me want to validate my own test results, if the ICP test providers would provide us with Gage R&R (Precision) and Accuracy data on their tests by parameter.
The information may be out there but I have not run onto it. I know that other have done this …Look Here
https://www.isixsigma.com/topic/gage-rr-for-icp-spectometer/ The closest I have seen is the article by Rich Ross & Dr. Chris Maupin in Skeptical Reefkeeping (Link Above) ….Although this is not a complete Gage R&R it gives some insight as to the
Accuracy of the test results (Table 1) related to each of the elements as well as some level of repeatability
(Precision) (3 samples from the same batch). When you take a close look at the Table 1 it is clear that, IMO, that some of the results would be considered accurate and fit for our use while other would not…Not wanting to through the baby out with the bath water…I will take the good ones and reject the inaccurate ones, but more data from all of the providers would be most helpful in helping us to maintain good water quality…IMO.
Sorry for the ramble!!
Rick
