The Uncertainty of an Estimate of the Bias ISO 5725-1 - Help

[01mercy]

Hello,

I'm in need of help for someone to look with me at a calculation to measure the values of the uncertainty (A) of an estimate of the bias of a measurement method at 95% probability with 5 laboratories (P=5) and 2,3 and 4 samples (n= 2,3 or 4) and with a reproducibility stdev - repeatability stdev factor of 1 (y=1).

The calculation in the standard is:

A = 1.96 SQRT( ( n * (y^2 - 1) + 1) /( y^2 * p * n ))

My results are
n=2 A=0.62
n=3 A=0.51
n=4 A=0.44

Results of the table in the standard are
n=2 A=0.76
n=3 A=0.72
n=4 A=0.69

I recalculated it over and over again... I can't find any mistake I make but I can't believe it's wrong in the standard.

Can anyone please re-check and if you have knowledge of such calculation please check if the formula is right stated, maybe it's a print mistake in the standard of the formula.

Thank you very much!

 

[dgriffith]

Are you dividing by y^2, then multiplying by * p * n, or are you dividing by the entire string? You might be missing a parenthesis?

 

[01mercy]

I divided the entire string.
I've written it down wrong here, just corrected it.
Thanks for noticing!

 

[dgriffith]

I get the same answer for n=4 as you (0.438), but I don't have access to the standard, nor have I ever seen it. Are you sure you didn't miss something in the book?
*Your math isn't wrong, but your equation or understanding of the parameters may need another look.

 

[01mercy]

iso 5725-1 first hit on google is the pdf download.

Page 9 clause 6.3.3.1 is the formula (formula 13)
Page 10 table 2 are the results of that formula

 

[Bill McNeese]

This won't help you much besides confirm that you have the same equation that is in the download ISO 5725 1994 standard you reference. I get the same as you when I do the calculation. I don't know if there has been an update to the standard since 1994 or there is something I don't understand. But the equation in the standard is straightforward. I could replicate the results in Table 1 of that standard.

 

[Bill McNeese]

Table 2 in the 1994 standard is not correct. I found this updated table 2 from a 1998 correction. It is attached. Your calculated numbers are correct as is the equation.

 

[dgriffith]

I have Parts 1, 2 and 3. Guess I should look for updates. Thanks for this one.

 

[01mercy]

Hi Bill,

Thank you very much. I didn't come across the update, I didn't scroll down far enough on the ISO site. shame on me.

Anyway, maybe still the thread is a good notice for people who look on this forum and experience the same issue with 5725