In multi-criteria decision-making and model evaluation, determining the weight of criteria is crucial. With the rapid development of information technology and the advent of the big data era, the need for complex problem analysis and decision-making has intensified. Traditional CRiteria Importance Through Intercriteria Correlation (CRITIC) methods rely on Pearson correlation, which may not adequately address nonlinearity in some scenarios. This study aims to refine the CRITIC method to better accommodate nonlinear relationships and enhance its robustness. We have developed a novel method named CRiteria Importance Through Intercriteria Dependence (CRITID), which leverages cutting-edge independence testing methods such as distance correlation among others. This approach enhances the assessment of intercriteria relationships. Upon application across diverse data distributions, the CRITID method has demonstrated enhanced rationality and robustness relative to the traditional CRITIC method. These improvements significantly benefit multi-criteria decision-making and model evaluation, providing a more accurate and dependable framework for analyzing complex datasets.