Additional to data collection relate and sample characteristics. Data were collected exclusively online, which may have limited both the diversity and representativeness of the participants. Furthermo

The gap

Consensus across the literature

Supporting evidence — 3 representative gaps

• Investigating Mathematics Teachers Technology Acceptance and Self-Efficacy for Technology Integration: A Structural Equation Modeling Approach (2026) · doi

  Additional to data collection relate and sample characteristics. Data were collected exclusively online, which may have limited both the diversity and representativeness of the participants. Furthermore, model fit indices were suboptimal, and the data did not fully meet the normality assumption, both of which restrict the generalizability of the findings. it For future research, is recommended to use measurement instruments that are culturally adapted and validated for the target population, to recruit a broader and more balanced sample, and to consider alternative structural modeling techniques, such as Partial Least Squares Structural Equation Modeling (PLS-SEM). Researchers may also explore whether treating Technological Complexity and Anxiety as separate factors or reverse-coding specific items within the Perceived Ease of Use dimension yields a more parsimonious and statistically robust model. 402 March 2026, Volume 18, Issue 3391-406Declarations Ethical Approval: Ethical Approval: This study was approved by the University institutional review board. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107(2), 238–246. https://doi.org/10.1037/0033-2909.107.2.238 to Participate:

  Keywords: structural sample model modeling ethical approval additional collection relate characteristics collected exclusively online limited diversity
• Technology affinity, digital tool use, and implications for health professions education: a cross-sectional study (2026) · doi

  This study benefits from a large and diverse sample, the use of a validated instrument (TAEG), and multivariate analysis techniques. However, several limitations should be noted. First, the study employed a cross-sectional self-report design, which limits causal interpretations and may be subject to socially desirable responding [14]. Second, the TAEG instrument used was the origi- nal version [19], which captures more established tech- nologies and may not fully reflect attitudes toward newer digital tools. Third, the number of independent variables included in the models was limited, hence it is possible that other factors such as organizational conditions or cultural attitudes—may play a significant role [2]. Longi- tudinal and mixed-methods studies are especially needed to unpack how contextual enablers—such as digital lead- ership, organizational culture, and time budgets—mod- erate the relationship between digital affinity and actual adoption behavior [2, 4, 28]. While the effect sizes were rather small, the high level of statistical significance— based on a large sample (N = 1211)—indicates a robust relationship. Even modest effects can carry practical relevance in large-scale or socially significant contexts, particularly when observed consistently across multiple domains [30, 31]. Fourth, the dichotomization of key variables, including digital tool usage and professional role, may have led to a loss of information and reduced the ability to detect more nuanced associations. In par- ticular, the binary operationalization of professional role (nursing vs. other health professions) grouped diverse healthcare occupations into a single comparison category Unkart et al. BMC Medical Education (2026) 26:782 Page 7 of 9 and therefore does not capture heterogeneity in roles, responsibilities, or digital exposure. As such, the study cannot assess whether technology affinity differs system- atically between specific professional groups. Although digital tool usage was initially assessed across differ- ent categories, the aggregation into a single composite measure does not allow differentiation between types, purposes, or complexity of technology use. For exam- ple, the measure does not distinguish between passive, entertainment-oriented use and active, task-oriented or professional applications. This limits the interpretability of the observed associations. The aggregation approach reflects a focus on overall digital exposure as a general behavioral indicator, rather than on specific technology domains. This aligns with the conceptualization of tech- nology affinity as a broad attitudinal construct rather than a domain-specific competence. Finally, the sample was based on an online survey, which could introduce self-selection bias, particularly among digitally confident individuals. Taken together, these limitations primarily affect the granularity—but not the overall direction—of the observed associations.

  Keywords: digital professional large sample role affinity rather observed associations technology specific diverse instrument taeg limitations
• The role of social media use motivations in university students’ adoption of AI-supported learning tools: The mediating effect of perceived usefulness (2026) · doi

  Several limitations should be acknowledged when interpreting the present findings. First, the sample was predominantly female (74.1%) and composed largely of first-year students (71.4%) from Kazakhstani universities, which may restrict the generalizability of the results to more balanced or internationally diverse populations. In addition to this the cross-sectional survey design precludes causal inference; longitudinal designs would be better suited to establish the temporal ordering of motivations, PU, and BI. The social connection/FOMO subscale of the MSMU demonstrated relatively low internal consistency (α = .645), which may have attenuated the estimated effects for this dimension and should be interpreted cautiously. The study did not include perceived ease of use, social influence, or digital literacy as control variables, all of which have been identified as relevant predictors in comparable technology adoption models (Alkhawaja et al., 2022; Yan et al., 2024), and their omission may constitute a source of confounding. Furthermore, all constructs were collected from a single respondent via a single online survey, raising the possibility of common method bias (CMB). Although procedural remedies were employed (anonymity, voluntary participation), statistical assessment of CMB or example, via Harman’s single-factor test was not conducted and represents a limitation. Finally, BI was measured rather than actual usage behavior, and the gap between intention and behavior is well-documented in the technology acceptance literature. CONCLUSIONS AND RECOMMENDATIONS The present study provides empirical evidence that university students’ motivations for using social media are meaningfully associated with their intention to adopt AI-powered learning tools, and that this relationship is partially mediated by PU. Grounded in the TAM and uses and gratifications theory, the findings demonstrate that social media motivations function as upstream antecedents of technology adoption not only through functional appraisal pathways but also through direct motivational processes. In particular, civic and advocacy motivations emerged as the strongest drivers in both direct and indirect pathways, underscoring the importance of information-oriented engagement dispositions for AI tool adoption in educational contexts. For university educators and instructional designers, these findings suggest that AI tool integration initiatives are likely to be most effective when they are framed within the information-seeking and collaborative knowledge-building motivations that students already exhibit in their social media use. Rather than simply introducing AI tools as productivity utilities, pedagogical approaches that position generative AI as an instrument for civic engagement, critical inquiry, and knowledge advocacy may

  Keywords: motivations social students technology adoption single media present first survey rather behavior intention university tools