When Performance Testing Is Not Feasible: A Methodological Perspective on Digital Skills Assessment in Higher Education

Emil Frashëri
Fan S. Noli University of Korça, Faculty of Economy,
Department of Management
Str: Shëtitorja Rilindasit, Korçë 7001, Albania
https://orcid.org/0009-0003-1062-8432
efrasheri2@unkorce.edu.al

Abstract
Digital skills are essential for participation in digital, data-driven, and AI-enabled societies. While performance-based assessments remain the gold standard, their financial, technological, and organizational demands often exceed the capacity of resource-constrained higher education institutions. This paper proposes a practical framework for digital skills assessment that combines validated self-report measures with second-generation structural equation modeling (SEM). Drawing on evidence from a research project conducted at an Albanian university, it shows how Exploratory Structural Equation Modeling (ESEM), Hierarchical Exploratory Structural Equation Modeling (H-ESEM), and bifactor-based SEM approaches can improve construct representation, reduce measurement bias, and strengthen the assessment of digital competence. Findings indicate that digital skills are multidimensional, hierarchical, and composed of interconnected functional and critical dimensions. Furthermore, digital skills profiling informed by latent profile analysis of bifactor-ESEM–derived factor scores facilitates evidence-based, targeted, and personalized curriculum development. The proposed framework offers a feasible alternative to performance testing while supporting evidence-based decision-making in under-resourced higher education institutions.

Keywords: Digital skills assessment, Digital competence, Second-generation structural equation modeling, Resource-constrained higher education, Digital skills profiling.

Article Type: Open Access

Full Text: PDF