HOBBY GRADE TEST KITS CAN OUTPERFORM ICP MEASUREMENTS…REALLY??

[Reefahholic]

I made sense to me after I thought about it, but I’m not a PhD level chemist like Christoph.

When we collect ICP samples, we only collect 48 hrs “after” the last water change to minimize issues similar to this. Although we don’t know for sure, so it might be good to check for any future experiments.

 

[MnFish1]

At the end of this discussion I think we all learned something. I hope to see some of you guys ICP testing soon.




I think so. Maybe send one like you previously did it, and another with NSW. See how much variability there is with iron in both samples.

For example - every time a sample on this testing was 'spiked' - it introduces error (or potential error) - This was. not taken Into account. The error that I see - is +- 20 percent on multiple measurements. Am I reading it incorrectly?

 

[Rick Mathew]

Actually - IMHO - hypotheses that seem 'easy enough to test' are often extremely difficult - unfortunately like the hypothesis you're trying to 'prove' here.

Yes...sometimes seemingly easy experiments turn out to be a challenge, but that is what experimental science is all about...Try---analize---try again...if you stay with it long enough you might get an answer

 

[Reefahholic]

For example - every time a sample on this testing was 'spiked' - it introduces error (or potential error) - This was. not taken Into account. The error that I see - is +- 20 percent on multiple measurements. Am I reading it incorrectly?

That could be the cause. We’ll leave that for another discussion. ;-) We’re all looking for good data so I’m not opposed to any testing.

 

[MnFish1]

Yes...sometimes seemingly easy experiments turn out to be a challenge, but that is what experimental science is all about...Try---analize---try again...if you stay with it long enough you might get an answer

Actually - I asked a specific question - the measurements in multiple occasions were +-20%. -am I incorrect>

 

[Rick Mathew]

Actually - I asked a specific question - the measurements in multiple occasions were +-20%. -am I incorrect>

Not sure which measurements you are referring to?

 

[MnFish1]

Not sure which measurements you are referring to?

From your spreadsheets posted above - lets take this one for an example (screen shot)

 

[MnFish1]

This is not the only one - the issue to me is that you're using (I think standard deviation - but I might be missing if its something else) - for the graphs - but - you do not have the same comparison to the ICP.

 

[MnFish1]

Not sure which measurements you are referring to?

Additionally there was no explanation (if I missed it I'm sorry) - as to where tester 3 went for this graph.

BTW - this is in no way a criticism of your hard work. However, I think some of the questions deserve an answer... Having Mean and Error bars for the 'testers' as compared to the ICP results - probably relates to cost - but it greatly takes away IMHO - any conclusion that anyone can make about the results

 

[Rick Mathew]

From your spreadsheets posted above - lets take this one for an example (screen shot)

If you are referring to the error bars on this chart they are based on multiple measurements for the chemical testers but we only got one measurement from the ICP vendors so no error bar could be assigned.

 

[Rick Mathew]

as to where tester 3 went for this graph.

Tester # 3 did not do a chemical test for copper

 

[MnFish1]

If you are referring to the error bars on this chart they are based on multiple measurements for the chemical testers but we only got one measurement from the ICP vendors so no error bar could be assigned.

Correct - so therefore they cannot be compared. And - BTW - of Course it would have cost a lot more to do 3 ICP tests than 1 - so I'm not trying to criticize you. Where did tester 3 go?

 

[MnFish1]

Tester # 3 did not do a chemical test for copper

Thanks! sorry my post crossed with yours.

 

[Rick Mathew]

Correct - so therefore they cannot be compared

Curious...What is your rationale for this statement?

 

[MnFish1]

Tester # 3 did not do a chemical test for copper

So I guess that rather than the somewhat click-bait title of the thread - (and no offense - if you didn't mean it that way) - the title could be - we tested 1/3 of the tests done on an ICP - including chemistry tests as compared to ICP - and ICP for the tests measured was (with poor statistical analysis) - pretty equivalent. However - people that want to use ICP to measure multiple elements - hobby tests are woefully inaccurate (since they do not exist)?

 

[Rick Mathew]

of Course it would have cost a lot more to do 3 ICP tests than 1

Yes this is true...In our last study we actually had 4 samples sent to each vendor and did a massive amount of statistics on the results to see their precision

 

[MnFish1]

Curious...What is your rationale for this statement?

Well - I don't see any statistical analysis of the data you presented - but - they can't be compared in the first place since you did 3 tests for some tests and 1 for the others. There is no way to judge statistical significance. You seem to be guesstimating - as to whether the ICP tests are equivalent or better/worse. I have no clue.

 

[MnFish1]

Yes this is true...In our last study we actually had 4 samples sent to each vendor and did a massive amount of statistics on the results to see their precision

Curious did you post these somewhere - not because I doubt you - but Im not sure that thats the way studies/statistics are supposed to work?

 

[Rick Mathew]

Curious did you post these somewhere - not because I doubt you - but Im not sure that thats the way studies/statistics are supposed to work?

How We Use ICP-OES Results Of Unknown Accuracy And Precision

A quantitative Method of Evaluating ICP-OES Results

www.reef2reef.com

 

[MnFish1]

How We Use ICP-OES Results Of Unknown Accuracy And Precision

A quantitative Method of Evaluating ICP-OES Results

www.reef2reef.com

Thanks - however - IMHO - you cannot use that data to compare the tests in this study. Thats my opinion. Others may differ