Page 1 of 1
Accuracy problems
Posted: Fri Aug 19, 2005 5:18 pm
by kst1
Hi! I have set up a headspace SPME method on our GC-MS. I am quantitating a compd. using an internal standard. I have a calibration curve with r-square= 0.993. Now, for validation purposes I am running samples of known concentrations. But the values after quantitation are all over the place. What could be the cause of this? Where do I start to look for this problem. The samples are being prepared the exact same way and everything....
Posted: Sun Aug 21, 2005 12:18 pm
by james little
I don't understand your problem totally.
You say you have a good calibration curve. Is the matrix of your calibration curve the same as that of the real samples.
If so, doesn't make any sense. If the sample matrix is different than your matrix to prepare the standard curve, need to make your calibration matrix more like your standard matrix.
Maybe you are getting some matrix effect due to interferences. Might be hard to tell if just doing selected ion monitoring instead of a full mass scan. THe other possibility would be different recovery efficiencies of your analyte in your matrix versus your standard matrix.
Posted: Tue Aug 23, 2005 8:54 pm
by kst1
Thanks for your reply. I am running the samples for validation in the same exact matrix (red wine). Yet, I am not getting very good results when I quantitate the samples with amts against the standard curve. Doesnt make sense and I am wondering if there is something I am doing wrong as far as quantitating the samples goes. How should I go about quantitating these samples with the standard curve generated? I am new to all this!!