Publicado en línea: 30 dic 2016
Páginas: 248 - 263
Recibido: 21 ago 2015
Aceptado: 27 jun 2016
DOI: https://doi.org/10.1515/foli-2016-0016
Palabras clave
© 2016 University of Szczecin
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.
A simple linear regression model is one of the pillars of classic econometrics. Despite the passage of time, it continues to raise interest both from the theoretical side as well as from the application side. One of the many fundamental questions in the model concerns determining derivative characteristics and studying the properties existing in their scope, referring to the first of these aspects. The literature of the subject provides several classic solutions in that regard. In the paper, a completely new design is proposed, based on the direct application of variance and its properties, resulting from the non-correlation of certain estimators with the mean, within the scope of which some fundamental dependencies of the model characteristics are obtained in a much more compact manner. The apparatus allows for a simple and uniform demonstration of multiple dependencies and fundamental properties in the model, and it does it in an intuitive manner. The results were obtained in a classic, traditional area, where everything, as it might seem, has already been thoroughly studied and discovered.