The effects of authentic and holistic leadership styles of the school managers on the positive organizational behavior: Application of artificial neural networks

The objective of the author is to predict the effects of authentic and holistic leadership styles of the managers on the positive organizational behavior. The sample group consists of class teachers working in the Turkish province Elazığ in the 2016-2017 school year. In this study, having been commissioned to determine results of the predictions regarding the effects of authentic and holistic leadership styles of the managers on the positive organizational behavior, the relational screening model has been used. To collect data, first of all the author has developed a "Holistic Leadership Scale" and the scope and the structural validity of the scale have been investigated. After also explanatory and confirmatory factor analysis has been made, the validity and the reliability of the scale has been accepted. As a second step, the Authentic Leadership Scale developed by Walumbwa, Avolio, Gardner, Wensing and Peterson (2008) and having been adapted into Turkish by Tabak, Polat, Coşar and Türköz (2012), and as a third step Psychological Capital (psycap) Scale developed by Luthans, Youssef and Avolio (2007) and having been adapted into Turkish by Çetin and Basım (2012) have been used to collect data. Thus, three different data collection tools have been used. In the study, independent sample t-test, one-way variance and correlation analysis were performed. Further, assuming that authentic and holistic leadership styles of the managers have an effect on the positive organizational behavior, logistical regression and artificial nerve networks analysis have been performed to predict the effects of authentic and holistic leadership styles of the managers on the positive organizational behavior. When authentic leadership and its subdimensions were analyzed in regard to two demographic data, one of them being gender and the other being seniority, no significant difference could be found in regard to gender. On the other hand, based on seniority it has been found out that the opinions of the teachers regarding the authentic leadership dimension, and the subdimensions associative transparency and self-awareness vary. It has been found out that the opinions of the teachers regarding positive organizational behavior and its subdimensions do not vary based on the gender and seniority variables. When the perception level of the teachers regarding the holistic leadership was investigated, it has been found out that in the holistic leadership dimension and in the management and teaching sub levels the opinions vary based on gender, while there is no significant difference in the support sublevel based on gender. On the other hand, it has been found out that the opinions regarding the holistic leadership dimension and its subdimensions the opinions vary based on seniority. The correlation analysis has shown that there is a significant association between the authentic and holistic leadership styles of the managers and the positive organizational behavior of the teachers. Also, the regression analysis has shown a significant association of all variables for the model (authentic leadership and associative transparency, internalized moral sentiments, unprejudiced evaluation, self-awareness subdimensions, as well as the holistic leadership and its subdimensions). When the prediction performances for all variables were evaluated, we found out that more successful classification results are achieved through the application of artificial nerve networks, and accordingly the prediction performance of the artificial nerve networks application is higher than the logistic regression model.