Network Determinants of Conflicts in Business Organization

This research is devoted to the investigation of network predictors task conflict in an organization. One such important predictor is employee centrality. There is mixed evidence regarding whether high centrality is desirable. For example, high betweenness centrality of a leader in the advice network was shown to be associated with higher conflict and lower team viability. Along with other aspects, my paper addresses such phenomena by an empirical study spanning three countries over four years. I use longitudinal data on one industrial organization, operating in Belarus, Poland, and Russia. The data were collected by the author via an online questionnaire, filled by the employees in 3 waves (June 2015, December 2017 and December 2018). Network data are based on employees’ nominations in response to questions on different working interactions, including reporting relationships, asking for advice, conflicts, and favorable attitudes. The data available also include employees’ position, department affiliation, seniority, peer-evaluated competences, social skills as well as demographics such as gender and age. I use these data to test the set of hypotheses about network factors of task conflict in an organization and to run an exploratory analysis in search of additional network predictors of conflict. The set of Machine Learning techniques is applied to solve the classification problem for the purpose of predicting a conflict link between a pair of employees. Demographic characteristics, measures of formal organizational position, peer-evaluated competencies and skills were used together with network-based characteristics of the involved employees to predict a conflict in a pair of nodes.