Network Psychometrics of the 10-Item Perceived Stress Scale among Patients with High Cardiovascular and Type 2 Diabetes Risk Using Exploratory Graph Analysis

Abstract

Background No studies have explored the internal structure of the 10-item Perceived Stress Scale in patients with high cardiovascular and diabetes risk. Objective We scrutinized the dimensionality of the scale in this patient group using exploratory graph analysis, a technique within the developing field of network psychometrics. Methods Analyses were conducted on 200 primary care patients. A bootstrap version of exploratory graph analysis assessed the stability of the dimensions based on structural consistency, item stability, and network loadings. Results Exploratory graph analysis revealed a 2-dimensional structure; structural consistency of the first dimension was high (0.863), whereas that for the second was low (0.667). Items belonging to the latter dimension did not cluster consistently with each other (ie, low item stability) and were not strongly associated with any particular dimension (ie, weak network loadings). Conclusion Exploratory graph analysis offers unique outputs, making it easy to assess the dimensional integrity of scales. Further research is warranted regarding the second dimension of the Perceived Stress Scale.