Answers are grounded in this paper and cite their sources.
Literature review analyzes factors influencing digital language tests' effectiveness, emphasizing student engagement and performance outcomes.
Given that digital language testing is increasingly embedded in university assessment systems, this review examines how technological, psychological, and contextual factors jointly shape students’ engagement, performance, and score interpretability. An expanded PRISMA 2020 search identified 1,878 records and retained 105 studies for synthesis. Although many studies interpret adoption through acceptance frameworks, the synthesis indicates that determinants extend beyond perceived convenience. Across the literature, determinants cluster into learner resources (digital literacy, technology-related self-efficacy, affect regulation); familiarity as a quality-dependent moderator; platform constraints and learner needs (diagnostic feedback, stability, accessibility); facilitating conditions (infrastructure, training, support services); and result demonstrability (transparent reporting and actionable feedback). Institutional support showed a strong positive effect on satisfaction, with an effect size of β = 0.458, p < .01. Additionally, institutional support functioned as a mediator between perceived usefulness (β = 0.366) and perceived ease of use (β = 0.198). In a university sample using a generative language assistant, usefulness (β = 0.644) had a stronger positive impact on user attitudes than ease of use (β = 0.301), while the model accounted for 81.1 per cent of the variation in user attitudes. Given that construct-irrelevant digital demands can depress performance, an anxiety-focused survey (N = 100) reported widespread worry about technical failures and heightened anxiety during online assessment. Overall, the findings indicate that platform design, preparation, and reporting should be equity-sensitive so that scores reflect language ability rather than digital access. Future research should determine how these clusters jointly predict achievement and the interpretation of scores across contexts.
No takes yet. Share an insight, caveat, or question.
Zhang et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: