Testowanie jednoczesne przy weryfikacji ocen parametrów strukturalnych liniowego modelu ekonometrycznego

In the process of verifying an econometric model, two stages can be identified. The first embraces “commonsensical” analysis of the results: whether they are in accordance with the theory of economy or observable results of experiments. The second consists of statistical analysis in which we test the statistical significance of this model both by using the global test and by testing each partial regression coefficient separately. Each of these analyses is usually conducted at the α level and the fact of multiple testing is disregarded. Disregarding multiple testing may result in bad decisions being made regarding the statistical significance of regression coefficients, which are in fact insignificant. Applying the procedures of multiple testing during the final stage of developing the econometric model used to test the significance of structural parameters of the linear econometric model allows us to determine whether the remaining structural parameters have been regarded as significant only because of multiple testing during the verification of their statistical significance.