The concentration of the analyte in the unknown sample must be calculated in the linear part of the standard curve because in this linear portion the instrumental signal is directly proportional to the concentration of the analyte in the unknown sample.