Chemistry PTP ststistics issues NADCAP recommended organiser

Please take a look at the image attached, what do you think to the distribution?

When 61 labs analyse a sample you would expect a normal distribution evenly spread around a the mean. But that is not what has happened here the "extreme values" are not evenly distributed - if the mean really is 0.04% then one would expect the same number of labs reporting higher as lower. But actually 6 labs have reported lower and only 2 labs higher. I would say that is evidence of a bias.

I think that the degree of rounding has reduced the resolution of the data, disguised the mean and the standard deviation is meaningless.

The same artifact is present to some degree on all elements analysed in the study but C is the worst example

Does it matter? We all passed and get the tick in the box for NADCAP.
But there are 2 points I think are important:

1. to maintain NADCAP we have to pass these PTPs regularly and investigate any failures, if an organisation has the power to say your results are wrong then the statistical methods and planning of the study should be beyond reproach - it aint

2. Many labs will go on to use the samples as equivalent to certified reference materials and the bias introduced by the rounding will be incorporated into their future analyses

I plan to complain to the supplier, but I raised the same issue with them last year, and although they accepted the argument; they have done exactly the same thing this year. I am also considering raising it with PRI/NADCAP. Would I be wasting my time, will anyone want to think about this and do anything?

I guess I am posting to ask if you agree with my observations and conclusions, whether its worth making a fuss and if so to whom?

 

That would make the organisers the swindling taylors and PRI the emperor. But if I cast myself as the small boy, will any body listen or will I just get a clip around the ear and told not to question my betters?