Title: Structural equations and machine learning: an approach to measuring political stability and citizen support in social networks

Authors: Edwin Arango Espinal; Carlos Fernando Osorio Andrade; Carlos Alberto Pastrana Arango

Addresses: Department of Administration and Organizations, Universidad del Valle, Caicedonia, Valle del Cauca 762540, Colombia ' Sinergia Group of the Instituto de Educación Técnica Profesional de Roldanillo (INTEP), Tuluá, Valle del Cauca 763021, Colombia ' Department of Administration and Organizations, Universidad del Valle, Cali, Valle del Cauca 760000, Colombia

Abstract: This study examines influence of political stability on digital citizen engagement across the 51 municipalities of the Colombian Coffee Cultural Landscape. Research employs a mixed-methods approach, combining content analysis to operationalise the dimensions of political stability, natural language processing to analyse comments on Facebook, and structural equation modelling to explore the relationships between variables. The findings reveal that political legitimacy and policy continuity are negatively associated with the volume of comments but positively associated with their valence. Political strength shows a positive relationship with both the volume and valence of the comments. This study offers original value by integrating governance concepts with social media analysis in a Latin American context, providing insights into how dimensions of political stability influence digital citizen engagement. The findings have implications for understanding the relationship between political stability and citizen engagement in the digital age.

Keywords: political stability; governance; social media; citizen engagement; political legitimacy; comment valence; comment volume; democracy; natural language processing; structural equation model.

DOI: 10.1504/IJEG.2025.145737

International Journal of Electronic Governance, 2025 Vol.17 No.1, pp.106 - 125

Received: 06 Sep 2024
Accepted: 19 Jan 2025
Published online: 18 Apr 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article