Multiplex network analysis of employee performance and employee social relationships

In human resourcemanagement, employeeperformance is strongly affected byboth formal and informal employee networks. Most previous research on employee performance has focused on monolayer networks that can represent only single categories of employee social relationships. We study employee performance by taking into account the entire multiplex structure of underlying employee social networks. We collect three datasets consisting of five different employee relationship categories in three firms, and predict employee performance using degree centrality and eigenvector centrality in a superimposed multiplex network (SMN) and an unfolded multiplex network (UMN). We use a quadratic assignment procedure (QAP) analysis and a regression analysis to demonstrate that the different categories of relationship are mutually embedded and that the strength of their impact on employee performance differs. We also use weighted/unweighted SMN/UMN to measure the predictive accuracy of this approach and find that employees with high centrality in a weighted UMN are more likely to perform well. Our results shed new light on how social structures affect employee performance. © 2017 Elsevier B.V. All rights reserved. * Corresponding author at: School of Economics and Management, Xidian University, Xian 710071, China. E-mail addresses:[email protected] (M. Cai), [email protected] (W. Wang). http://dx.doi.org/10.1016/j.physa.2017.08.008 0378-4371/© 2017 Elsevier B.V. All rights reserved. 2 M. Cai et al. / Physica A 490 (2018) 1–12