The method introduced in the article allows for sensitivity analysis in various types of models, such as linear and nonlinear ones, without requiring extra calculations. By leveraging the duality property of mathematical programming, the method converts parameters into artificial variables to easily determine their impact on the model. This approach is particularly useful for regression models and optimization problems, including least squares and Weibull distribution estimation from censored data. The method is not only versatile but also computationally efficient, making it a valuable tool for researchers in different fields.