lod and loq again

[n-cute]

I am wondering how to calculate lod and loq. Is it possible to use S/N ratio to do this?

[grreddy]

I am wondering how to calculate lod and loq. Is it possible to use S/N ratio to do this?

You can use S/N Ratio for calculating the LOD and LOQ the method is S/N should be 10 and for LOQ it is 3

[n-cute]

Thanks for your reply. Are there different approaches to this? e.g. estimating noise value and transforming it to concentration, multiplied by 3 or 10? (yep, propably it's rookie question )

[grreddy]

Is there you just check agilent web site link you will get

[Mark Tracy]

Basically, yes. Measure the noise of a matrix blank over the time-interval of your peak. Most people use the peak-to-peak method for noise measurement because that most nearly matches the behavior of your peak integration software. Divide that number by the slope of the calibration line and multiply by 3 for LOD or by 10 for LOQ.

[n-cute]

Thanks for advices. Can I know, what method/s You use?

[Mark Tracy]

I usually use the S/N method, when it is my choice. Sometimes the regression method is better. As an EPA auditor once told our class, pick a method that you can statistically defend, write it into your SOP, and be consistent in using it. And, no you don't have to use the same LOD method for different projects.

[n-cute]

Thanks again for Your helpfull advices. It's good what this forum exist (forum=webpage + users)