Turning to comparability across groups, invariance results were mixed across assessments. In PISA, the bidimensional model achieved configural, metric, and scalar invariance across countries, providing strong evidence that the construct is comparable across participating education systems (supporting H2). In ICILS, the model reached configural and metric invariance but did not meet scalar invariance criteria across countries, suggesting that cross country mean differences should be interpreted cautiously and reinforcing the importance of explicitly evaluating between country comparability before drawing substantive conclusions (partial support for H2). These results indicate that even when the same two-dimensional structure is recoverable, the degree of cross national comparability may depend on assessment specific design features and item functioning. In contrast, gender invariance results were consistently supportive. Both assessments achieved scalar invariance across gender, indicating that General and Specialized DSE can be compared meaningfully between boys and girls within each dataset (supporting H3). This finding strengthens the validity of gender comparisons in DSE and suggests that observed gender gaps in DSE are unlikely to be driven by measurement non-equivalence, at least within the tested frameworks.

while others struggle(Rohatgi et al., 2016; Scherer & Siddiq, 2015).

Ismael Aguayo2,

Daniel Miranda1,

supporting H4 in a cross national comparability.

Gender differences were clearly dimensional specific: girls tended to report slightly higher General DSE, whereas boys reported substantially higher Specialized DSE across most education systems. This pattern indicates that gender disparities depend on the type of digital task domain being evaluated (supporting H4).

[European Commission Joint Research Centre (2022); Ulfert-Blank & Schmidt (2022)]

"specialized"

or computational

In sharp contrast

“logical solution”, “programming”). Additionally, it showed a positive residual correlation with “Privacy settings”

The high covariance indicates that students perceive multimedia creation and web development as a unified dimension of ‘content creation’, a finding that aligns with the DigComp framework (‘3.1 Develop Digital Content’).

indicators suggesting

Responses as “I don’t know” was threated as missing values.

Official subset of data reach around 393,000 students, nested on 52 countries (maily OECD).

[Hristov et al. (2026)]

may matter. while self-efficacy

within a single line of inquiry (Aesaert & van Braak, 2014; Papastergiou, 2010). Within

In turn, the literature consistently reports that students with low expectations of specialized self-efficacy sometimes score higher on standardized tests of digital skills

The dimensions of the DSE are not only relevant to contrast for theoretical reasons, but also because consideration of this approach has an impact on the distribution among groups

The problem is that recent definitions of Digital Competence are no longer framed within a bidimensional approach to self-efficacy with technologies.

we aim to clarify whether differences in DSE are consistent across contexts or instead a product of how assessments operationalize the construct.

(ulfert-blank_assessing_2022?) suggests to work with a unified construct denominated Digital Self-efficacy (hereinafter DSE) to reach a high-level research on this issue. Considering the gaps and inconsistencies in previous measurements, (ulfert-blank_assessing_2022?) points out that DSE construct have to

. PISA has a one-dimensional approach, as can be inferred from a single battery consisting of 10 items