PURPOSE: To correct the linear and nonlinear errors of a detector so as to improve the measurement accuracy of the detector by measuring the relative positional deviation between each diffraction grating pair of a plurality of diffraction grading pairs for correction and correcting the relative positional deviation based on the relational approximation between the measured and actual positional deviations between each diffraction grating pair.
CONSTITUTION: A table 5 on which a diffraction grating group 4 having a known relative positional deviation used for correction is formed is positioned at the same height as that of a wafer 3 to be measured and the positional deviations of the diffraction grating pairs are successively measured. When a graph indicating the relation between the actual (known) and measured positional deviations is plotted and expressed by a linear approximation, X/Pb=F(Xm/Pb) (where, Xm, X, Pb, and F respectively represent the measured and actual positional deviations, designed pitch value on a reticle, and constant) is obtained. Therefore, when the constant F is found from the measured results and measured values are corrected by multiplying the measured results of a diffraction grating pair having an unknown relative positional deviation by the constant F, the nonlinear and linear errors of the electrical system of a detector can be corrected.
MORI TETSUZO
CHITOKU KOICHI
