Schwarz Information Criterion ( \({\text{SIC}}\) ) serves as a measure of goodness-of-fit, which in other versions of the formulation contains the coefficient of determination. This formula provides a basis for linking the adjusted signal-to-noise ratio (the adjusted SNR) to the \({\text{SIC}}\) , enabling modifications to the \({\text{SIC}}\) framework in this research. The results of simulation studies on autoregressive (AR) order determination and regression model show that the probability of the candidate model fitting the true model based on the modified \({\text{SIC}}\) ( \({\text{SIC}}^{*}\) and \({\text{SIC}}^{{**}}\) ) is greater than the probability of the candidate model fitting the true model based on the classical \({\text{SIC}}\) . The results of this research demonstrate that \({\text{SIC}}^{*}\) and \({\text{SIC}}^{{**}}\) offer an improves ability to identify the best model among candidate models.