Connecting the Dots: Exploring Psychological Network Analysis as a Tool for Analyzing Organizational Survey Data

Abstract

Organizations allocate considerable resources in surveys aimed at assessing how employees perceive certain job aspects. These perceptions are often modeled as latent constructs (e.g., job satisfaction) measured by multiple indicators. This approach, although useful, has several drawbacks such as a strong reliance on local independence and a lower performance in exploratory contexts with many variables. In this paper, we introduce psychological network analysis (PNA) as a novel method to examine organizational surveys. It is first argued how the network approach allows studying the complex patterns of attitudes, perceptions, and behaviors that make up an organizational survey by modeling them as elements in an interconnected system. Next, two empirical demonstrations are presented showcasing features of this technique using two datasets. The first demonstration relies on original organizational survey data (N = 4270) to construct a network of attitudes and behaviors related to innovative work behavior. In the second demonstration, drawing on archival leadership data from an organization (N = 337), the focus lies on comparing structural properties of leadership attitude networks between subsamples of supervisors and non-supervisors. We conclude this paper by discussing how PNA constitutes a promising avenue for researching organizational phenomena which typically constitute a set of interconnected elements.