K-Means Clustering and Predictive Dependence of Personality Traits in Dark Triad Model

Machiavellianism, Narcissism and Psychopathy are the three traits expounded by the Dark Triad Model of Personality. These building blocks of personality cover a large and interconnected spectrum of personality. The use of Machine Learning can help in analysing largesized databases to draw meaningful comparisons. In this paper, we have used K-Means Clustering to investigate dark personality clustering in a very large database comprising 18192 respondents. The data was normalised and then subjected to unsupervised machine learning to classify it into four separate clusters. The data was clustered evenly among the four clusters. To evaluate the interdependence of the personality traits intraclass (0.537) and Pearson's correlations were computed. Two of the SD3 traits were used to regress on the third and the process was used for all of them alternatively. Pearson's correlation between Machiavellianism and Psychopathy was highest at 0.721 and the same was corroborated by the regression equations between the two traits. Narcissism showed equal correlations with Machiavellianism (0.475) and Psychopathy (0.453). Regression equation revealed that Machiavellianism was most significantly affected by Psychopathy and vice versa. Regression results for Machiavellianism and Psychopathy indicated fair predictive capabilities with R2 values of 0.547 and 0.535 respectively