Doğrusal olmayan kanonik korelasyon analizi ile liderlik vasıflarını etkileyen etmenlerin incelenmesi

Abstract

Using the categorical data, especially in the social science stuties, supported a rapid development of analysis techniques called Nonlinear Multivariate Analysis.These techniques are Correspondence Analysis, Homojenity Analysis, Nonlinear Principle Component and Nonlinear Canonical Correlation Analysis. The aim of the project is to find out the factors that effect the leadership features with the help of a studyon Nonlinear Canonical Correlation Analysis. Having aimed the project, first it was poused on scaling and quantification, next nonlinear canonical correlation was briefly studied. Then linear and non linear canonical correlation analysis was studied in detail and finally, the level of military and demographic formation effects on leadership features was tried to be determined with an inquiry just graduated ensigns The analyses were applied in for dimension, later the dimension amant was decreased to three and two, and finally the resultswere compered. The canonical correlation Analysis with the three groups was applied in two dimension. In conclusion, the leadership features are highly motivated for those who were admitted from the Naval High School, and for those with a mother who has low educational backgraund. The increase of the amount of the siblings in the family alsoaffected the leadership features positively. Since the inquries do not include quantative data, the hypothesis could verified. As a result the inquiry data were studied with the Nonlinear Cannonical Correlation and the opportunity was given to discuss the meaningful results that were shawn by graphics.