Modified Ramsey RESET in Combined Truncated Spline–Fourier Nonparametric Path Analysis on Waste Management Behavior

Nonparametric path analysis is a statistical approach that does not require the functional form of relationships between variables to be known a priori. Classical path analysis assumes linearity, which can be tested using the Ramsey Regression Specification Error Test (RESET). If the linearity test indicates that the relationships between variables are nonlinear, a nonparametric model can be applied. The purpose of this study is to develop a modified Ramsey RESET to identify nonparametric relationships modeled using truncated spline and Fourier series. The modified Ramsey RESET algorithm was successfully implemented to detect the optimal functional form of the nonparametric truncated spline and Fourier series and was subsequently applied to behavioral data on waste management practices. Furthermore, this study proposes an estimator for a hybrid nonparametric path model combining truncated spline and Fourier series approaches. The analysis results reveal that the best model integrates truncated spline with one and two knot points and a Fourier series with one oscillation. The model achieved an adjusted coefficient of determination of 0.956, indicating that it explains 95.6% of the variation in the Behavior of Transforming Waste into Economic Value, while the remaining 4.4% is explained by other unobserved factors outside the model.