Restoring the structure: A modular analysis of ego-driven organizational networks

TL;DR

This paper introduces a privacy-preserving method for organizational network analysis that uses ego-networks without PII to reveal large-scale organizational structures.

Contribution

It proposes a novel approach to generate organizational networks from anonymous ego-networks, addressing privacy concerns in ONA.

Findings

The method accurately reflects known organizational structures.

It enables large-scale analysis using only anonymous data.

The approach mitigates privacy risks in organizational studies.

Abstract

Organizational network analysis (ONA) is a method for studying interactions within formal organizations. The utility of ONA has grown substantially over the years as means to analyze the relationships developed within and between teams, departments, and other organizational units. The mapping and quantifying of these relationships have been shown to provide insight into the exchange of information and resources, the building of social capital, and the spread of culture within and between organizations. However, the ethical concerns regarding personally identifiable information (PII) that exist for traditional social science research are made more pertinent in ONA, as the relational nature of the network may leave participants open to identification by organization management. To address this, we propose a method of generating a network of organizational groups (e.g. units, departments, teams) through the projection of ego-networks absent of PII. We validate this method through modular analysis of the resulting networks and compare the identified structure to a known structure of the organization. The methodology lays a foundation for performing ONA that needs only anonymous ego-centric data to identify large-scale aspects of organizational structures.