15

Recent developments in data accessibility, such as the promotion of open data policies or the growing use of big data in organizations, are raising important methodological issues for HRM research. Management scholars, accustomed to working with small samples, face a major challenge when trying to exploit opportunities offered by big data or secondary data from large-scale surveys. One possible way to face this challenge is via a particular type of network analysis, allowing the visualization of relationships between a large set of variables. Early applications of this innovative approach emerged in the field of clinical psychology and now provide powerful and informative visualization tools, encouraging the development of an active area of research known as network psychometrics. This approach renews the study of measurement scales for attitudes and the evaluation of relationships between psychological variables, by relying in particular on parsimonious graphs (regularized Gaussian graphical models). This paper aims to demonstrate that this approach is promising when transposed to the field of organizational behavior. To illustrate the stakes of this adaptation, a French version of a recent measure of workplace commitment proposed by Klein et al. (2014) is examined, along with a tutorial presenting analytical and graphical tools developed in the open-source software environment R.