Calibration curves least-squares evaluation

When the uncertainty is negligible in comparison with S, the estimations by Eq. (7) do not differ from the estimation used in the ordinary least squares technique. However, the uncertainty increase will influence the result of such estimation. Theoretically it may even happen that o > S, and Eq. (7) lead to an absurd result. In such cases p]=°° is accepted and no absurd results will be obtained [6]. This influence is very important not only for determination of the analyte concentration X0 corresponding to the response Y0 by the calibration curve, but also for the correct uncertainty evaluation in the determination result. 

Under conditions of excess the reaction is pseudo-first-order in SCN . The curves in Figure 29-6a show the increase in absorbance due to the appearance of FefSCN) " versus time following the rapid mixing of 0.100 M Fe- " with various concentrations of SCN at pH 2. Since the concentration of Fe(SCN)-" is related to the absorbance by Beer s law, the experimental data can be used directly without conversion to concentration. Thus, the change in absorbance AT between times r, and f2 is computed and plotted versus [SCN jy, as in Figure 29-6b. Unknown concentrations are then determined by evaluating AA under the same experimental conditions and obtaining the concentration of thiocyanate ion from the calibration curve or the least-squares equation. 

Initial Calibration Verification (ICV) Standard prepared in a similar manner as for the working calibration standards, however, at a different concentration than any of the working standards. Used to evaluate the precision and accuracy for an interpolation of the least squares regression of the calibration curve. This standard is run immediately after the calibration of the working standards. 

---