I am new in gc and try to understand basic concepts.
My standard curve is a straight line within a certain range of concentration. From a definite concentration, it drops. Is this concentration the upper control limit???
Having a linear curve, is it a sign of "good results" or good sensitivity?
I have read somewhere that it is good to plot the response factor vs concentration and one should obtain an horizontal line. How do I calculate it and what is its meaning???
Thank you for your help
By MK on Monday, July 19, 2004 - 09:33 am:
Yes, it is your linear range. It's normal that the line starts to bend at high concentrations. You can dilute the sample if you don't hit the linear range.
Linear curve means you can suucesfully use linear quantitation. The slope of the curve tells you about sensitivity.
commonly a correlation coefficient of the linear regression is used to assess the linearity of the method. However this is usually not enough and response factor vs concentration is employed as an additional check. Your method is truly linear if you get a straight line.
By carla on Sunday, August 15, 2004 - 02:08 pm:
But what to do, if the calibration curve does not pass really the origin?
What happend, when it crosses the amount/concentration-axis and what if the curve crosses the response-axis?
Maybe, you dont have to dilute the sample if you use split mode, you can try to increase the split ratio.
What standard deviation is acceptable for the response factor vs amount/concentration, what do you get with your methods?
<1%? <3%? <5%?
By Travis on Friday, August 20, 2004 - 03:21 am:
I have the same question, I have my RSD values of Area Ratios of the sample and internal standard < 3%, R2= 0.956. I do injections manually on a split mode. Is it always mandatory to have 0.99 for R2 to represent the linearity?