Orthogonal signal correction (OSC) is a pre-processing technique used for correction of instrumental drift, bias and scatter in near infrared spectra. OSC separates the variation into orthogonal factors, where the factors contain the variation within the spectral data matrix that is not correlated with the analyte vector data. The aim of this study is to investigate different orthogonal factor selection methods, and identify effective orthogonal factor selection methods, which will enhance the performance of the OSC routine for quantitative analysis of near infrared spectra. In order for factor selection methods to be implemented on OSC, an amendment to the OSC algorithm is made, a binarized weighting matrix is applied to the OSC factors, which are used to generate a signal corrected prediction data set. The amended algorithm is termed weighted orthogonal signal correction (WOSC). Optimization of the binarized weighting matrix for appropriate selection of OSC factors is a challenging problem. The approach taken in this research is to identify promising heuristic techniques for the purpose of optimizing the weighting matrix. The use of factor optimization methods and strategies provide a more intelligent method of using the OSC algorithm. The data set used was potato crisp near infrared spectra. The potato crisps were not crushed or ground to produce a spectral dataset with scattered spectra. The optimization strategies tested were a genetic algorithm, hill climbing, feature selection, stepwise selection, and full spectrum modeling. Using the different selection strategies, different combinations of OSC, WOSC factors and spectral predictors were selected. Partial least squares regression was undertaken to form calibration models for the OSC and WOSC pre-treated data sets and the cross validated standard error was used as a measure of model performance. It was found that the WOSC algorithm combined with factor selection methods produced better cross validated standard errors relative to OSC pretreated data. This result suggests that WOSC may have some potential in automatic inspection applications using near infrared spectroscopy.