Purpose: The research aims to identify statistical, econometric, and economic indicators to infer the problem of spurious regression. Theoretical framework: The research used a descriptive and econometric approach to identify the dimensions of this problem and the most likely causes of its emergence. Design/Methodology/Approach: The research estimated four equations, the first two are simple linear regressions (y, x1) and (y, x2), the third is a multiple linear regression equation (y, x1, x2), and the fourth is a double logarithmic regression equation for multiple linear regression. The results of the four equations were compared to assess the standard methods and explore spurious regression. Findings: The research concluded with several findings, one of which is that a high value of the coefficient of determination (R2) in the absence of significance of the parameters is an indication of the problem of multicollinearity, and hence the possibility of spurious regression. Research, Practical & Social implications: Spurious regression can lead to misleading conclusions about the relationships between economic variables, this can have negative consequences for economic policymaking and other decision-making processes. The research on spurious regression can help to improve the accuracy and reliability of economic analysis, which can benefit society as a whole. Originality/Value: The paper could provide recommendations for how to avoid spurious regression in econometric models of imports. This could include recommendations for data collection, data preparation, and model specification.
Read full abstract