Please use this identifier to cite or link to this item: http://hdl.handle.net/10419/205081
Authors: 
Merce, Emillian
Drăghici, Manea
Berca, Mihai
Merce, Cristian C.
Necula, Raluca-Alexandra
Year of Publication: 
2018
Citation: 
[Title:] Agrarian Economy and Rural Development - Realities and Perspectives for Romania. 9th Edition of the International Symposium, November 2018, Bucharest [Pages:] 10-15
Abstract: 
The authors point out that autocorrelation is an accidental statistical phenomenon. The cause of the occurrence of collinearity is the incomplete data base and that ideally, the elimination of the occurrence of autocorrelation is achieved by knowing the effect stage for all possible combinations of the variants of the factors involved. In many cases, the practical realization of such a desideratum is impossible. Such a difficulty is especially specific to statistical processing in the economic, social and psychological field. Neither multi-factorial experiments of agrobiological nature are not avoided by such difficulties. Consequently, to the researcher remain at his disposal methods of distributing collinearity on influence factors using methods based on the calculation of partial correlation coefficients (Merce E., 1986; Moineagu C., 1974). With obvious computing facilities, compared to the evoked methods, the authors suggest using an original method based on the principle of proportional distribution of autocorrelation with the proportion of simple determinations, following the next six steps, the last four are solved instantly after the first two steps have been solved: 1. The calculation of the multiple correlation coefficient and simple correlation coefficients using the Regression function of the Data Analysis component of Microsoft Excel Program; 2. The recording of the multiple correlation coefficient and of the simple correlation coefficients; 3. The simple determination coefficients and the multiple determination coefficient calculation; 4. The sum of the simple determination coefficients; 5. The calculation of the simple determinations proportions, considering their sum equal to 100; 6. The determination of each factor influence, as a product of multiple determination and the proportion of simple determinations.
Subjects: 
autocorrelation
distribution of autocorrelation by factors
method and program
JEL: 
C40
Document Type: 
Conference Paper

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.