**3.4 Standardization of the analytical method for cadmium determination**

Statistical quality parameters must be evaluated in order to standardize the analytical method, as described below.

## *3.4.1 Linearity*

To examine linearity, a calibration curve needs to be prepared in a concentration range including at least six different cadmium concentrations between 0.1 and 1.0 μg Cd/L. This should be analyzed using the atomic absorption equipment, taking four absorbance readings for each cadmium standard. The calibration curve should be plotted (Absorbance vs. concentration) and the correlation coefficient and slope should be statistically evaluated, through statistical treatment of the parametric data obtained for each concentration level [54, 70].

A regression analysis for the model Y = β0+ β1X is required to confirm whether the degree of relationship is significantly linear between the two variables (absorbance and concentration). **Table 3** illustrates the coefficient that defines the slope of the regression line and the t-statistics for cadmium, making it possible to contrast the null hypothesis (Ho) that the slope has a value of zero. According to the results, the calculated statistics are greater than the tabulated ones, confirming that the slopes differ significantly from zero and therefore, the absorbance is significantly correlated to the concentration [54, 70].

The one-way ANOVA was applied to confirm the association between the two variables (absorbance and concentration) by means of a linear regression for cadmium. The results showed that the calculated F (6119.282) was higher than the tabulated F (4.080) and therefore, the null hypothesis raised is rejected (Ho = There is no significant linear relationship between absorbance and concentration) [54, 70].
