Analyte calibration and

Recovery tests or use of internal standards for pre-treatment and use of reference materials for analytical calibration and traceability [5]. 

A variety of criteria can be used to fit a model to a set of data points. Consequently there is no single best straight line through a set of points. What is best depends on how the data were collected and the exact purpose of the regression. Among the possible criteria, the common criterion of unweighted least squares (LS), also known as the ordinary least squares (OLS) fit, is by far the most popular one for analytical calibration and will be considered here. 

---