Validation of a Tool to Measure the Digital Competence of Students from Universidad Latina

Abstract

Digital competence (DC) is essential in higher education, supporting academic, professional, and cultural development. This study aimed to validate the COMPDIG-PED instrument for assessing digital competence among students at Universidad Latina.  Data were collected using non-probabilistic convenience sampling during the second quarter of 2024.  Reliability analysis revealed satisfactory internal consistency, with Cronbach´s alpha (α = .70). However, the distribution of item difficulty highlighted a significant imbalance, as 36.51% of items were classified as very difficult, while only 12.70% fell into the medium-difficulty range.  Participants scored between 12 and 41 (M = 31.52, SD = 5.05) out of a maximum of 63 points. These results highlighted the need to recalibrate item difficulty to enhance the instrument´s diagnostic value. Confirmatory factor analysis further raised concerns regarding the instrument´s unidimensionality. The Kaiser-Meyer-Olkin (KMO) value was low (2.0), and Bartlett´s test was significant (p< .001), indicating weak inter-item correlations. Additionally, low alpha values (0.13 – 0.55) within specific factors suggested issues with item alignment and clarity.  These findings emphasize the need for refinements to the COMPDIG-PEG instrument to improve its design, structure, and reliability, ensuring it provides a comprehensive assessment of digital competences in higher education contexts.