Replicates of calibration curve lowest standard

[Star Stella]

Hi everyone,

I have a question regarding calibration curve lowest standard. Do you prepare lowest and highest standards in duplicate?
In this case are you using mean value or the first preparation value in the regression?

Thanks in advance!

[tangaloomaflyer]

nope, we don't do this.

[tom jupille]

In principle, you should use *all* the data points.

[James_Ball]

Depends on what work you are doing. If you are running under EPA methods it isn't allowed, you only use one standard determination per level, no averaging. I do often prepare them in duplicate and use them to QC the curve afterwards, since if the curve is good and instrument is behaving properly you should be able to quantify them back accurately. Lowest should be able to hit +/- 50% and the high +/- 20% or much less.