Understanding accuracy percentages of test kits and/or equipment

[oldsaint]

I remember seeing Boomer explain this once before but I'm not sure where I ran across it. IIRC, it's not as simple as it seems.
If my test kit has an "Accuracy 4% ±0.04 ppm (mg/L) of reading 0.04 ppm (mg/L)" does that mean at a reading of .05 the accuracy is +/- .002 ppm or the reading could be some where between .09-.01 ppm?

I think I remember something about the accuracy changes proportionally to the reading.

 

[spieszak]

hmmm... I thought accuracy was constant when within the range of the test kit. I'll be tagging along to see if (when) I'm totally wrong.

 

[Mechanic1]

Which is more important accuracy or consistency??

 

[spieszak]

depends on how inaccurate the consistency is?
i prefer consistent, but for new(er) folks, accuracy tends to be more important. i can read my tank a bit now without having to test. when i couldn't, i needed accuracy to diagnose... now i want consistency since i have tank indicators. tank tells me somethings off, i just need to know which of my numbers test wrong (compared to my consistent results), not how accurate the number really is..

 

[Mechanic1]

That is my thinking also.
My corals can indicate when something's not right.

 

[Boomer]

Oldsaint

That does not look like it is written right. Here is a known accuracy for a meter


Iodine ±0.1 mg/L ±5% of reading

It means you have a reading, where you take ± 0.1 mg / l and ±5% of the reading.

If the reading is 10 ppm then;

10 mg / l - 0.1 mg/ l = 9.9 mg/l

9.9 x 5 % = 0.495 gives

9.405

So, at a reading of 10 ppm they are guaranteeing the actual level is somewhere between 9.405 - 10.405 ppm, you do not know where. So, round that off and you have an Accuracy @ 10 ppm of ± 0.4 mg/ l.

If you were to work out 3 mg/ l

3 - 0.1 = 2.9

2.9 x 5% = 0.145 gives

2.755

Accuracy @ 3 mg/ l ± 0.245 mg/ l or ± ~ 0.2 mg/l

So, your actual level is somewhere between 2.8 - 3.2 mg/ l, you do not know where.

 

[oldsaint]

It's not written correctly. I see that now. Not sure where I ran across it but it had me confused.

I tested a fellow reefer's tank with a HI-97713. He used his 713 checker a few days later and got a higher reading. I went to see if the readings were within range and got that some where.

I have both meters with me and looking through I see that both have the same accuracy ±0.04 (mg/L) ±4% of reading(mg/L) and see how the math works.

Thank you very much.

 

[Mechanic1]

:doh::doh:

Oldsaint

That does not look like it is written right. Here is a known accuracy for a meter


Iodine ±0.1 mg/L ±5% of reading

It means you have a reading, where you take ± 0.1 mg / l and ±5% of the reading.

If the reading is 10 ppm then;

10 mg / l - 0.1 mg/ l = 9.9 mg/l

9.9 x 5 % = 0.495 gives

9.405

So, at a reading of 10 ppm they are guaranteeing the actual level is somewhere between 9.405 - 10.405 ppm, you do not know where. So, round that off and you have an Accuracy @ 10 ppm of ± 0.4 mg/ l.

If you were to work out 3 mg/ l

3 - 0.1 = 2.9

2.9 x 5% = 0.145 gives

2.755

Accuracy @ 3 mg/ l ± 0.245 mg/ l or ± ~ 0.2 mg/l

So, your actual level is somewhere between 2.8 - 3.2 mg/ l, you do not know where.

Wow to much math