Purpose In digital learning environments, learners need to regulate their learning, yet many do not engage in effective self-regulated learning strategies. Prompts can scaffold these processes, but their effectiveness may depend on their specificity and alignment with learners’ ongoing learning behavior. The purpose of this study is to address the limited evidence on the effects of generic, specific, and behavior-adaptive prompts on learning processes and outcomes. Design/methodology/approach This study (n = 64) uses an experimental between-subjects design investigating generic (GP), specific (SP) and adaptive specific (AP) and no prompts with regard to learning performance, perceived learning support and navigation behavior. Findings No significant group differences were found for declarative knowledge, whereas significant differences emerged for knowledge transfer, perceived learning support and navigation behavior. Independently of the experimental condition, higher transfer performance was associated with navigation that deviated more strongly from the course structure. Research limitations/implications Future research should investigate learners’ interactions with and reactions to prompts, for example through eye-tracking, screen-recording, and brief self-report measures. Originality/value By comparing fixed generic, fixed specific, adaptive specific and no prompting, this study considers prompt specificity and behavior-contingent adaptivity together. It also combines learning performance and perceived learning support with trace-based navigation behavior. Fixed generic prompts were less effective for knowledge transfer than adaptive specific prompts or no prompting, while adaptive specific prompts did not outperform fixed specific prompts or the control group. The findings therefore challenge the assumption that providing prompts or increasing their adaptivity is inherently beneficial and highlight the importance of prompt specificity, timing and fit with learners’ ongoing behavior.
Purpose The rapid integration of Conversational Generative Artificial Intelligence (GenAI) into higher education has sparked intense pedagogical debates regarding its impact on student cognition. This study aims to deconstruct the micro-level interactive processes through which university students engage with large language models (LLMs), mapping the systemic socio-technical mechanisms that shape, extend, or restrict the development of independent critical thinking skills. Design/methodology/approach Adopting an exploratory qualitative research paradigm, this study used a combination of purposive and maximum variation sampling strategies to recruit n = 55 undergraduate and postgraduate students across diverse disciplinary tracks (STEM, medicine, humanities and social sciences) at a public university. Data collection was executed through semi-structured depth interviews incorporating stimulated recall methodologies and live interaction histories. The qualitative texts were analyzed using Braun and Clarke’s inductive thematic analysis methodology guided by Grounded Theory coding protocols via NVivo, augmented by cross-dimensional matrix queries across student subgroups. Findings The analysis generated an integrated socio-technical framework comprised of four core themes. The Paradox of Cognitive Offloading vs Extension, where automated text consolidation yields significant efficiency dividends but risks fostering cognitive inertia and domain-specific skill atrophy, particularly in technical code and math execution. The Awakening of Algorithmic Discernment through Epistemic Vigilance, revealing that model imperfections, flawed logic and fabricated bibliographic citations serve as unexpected cognitive catalysts that disrupt passive trust and force active multisource validation loops. The Regulatory Role of Individual Prompt Literacy, distinguishing low-literacy single-turn commands from advanced interactive conditioning workflows that deconstruct complex problems; and the Scaffold Effect of Process-Oriented Pedagogical Interventions, demonstrating that instructor-imposed structural constraints create “desirable difficulties” that effectively preserve human cognitive agency. Research limitations/implications The research relies on qualitative self-reported reflections during depth interviews and represents a cross-sectional snapshot of rapid technological evolution. Future research should integrate objective backend telemetry data with multi-semester longitudinal tracking. Practical implications Grounded directly in empirical student interaction data, the paper provides an actionable blueprint for higher education governance, advocating for a transition from unenforceable prohibition policies toward structured “Critical Integration.” Key recommendations include rewriting curriculum assignments to evaluate process-oriented modification logs rather than final written products, embedding formal prompt literacy into core curricula and training students to deploy multisource triangulation protocols against digital hallucinations. Originality/value Moving beyond descriptive behavioral intention surveys and binary debates labeling GenAI as either a destructive shortcut or a perfect accelerator, this study opens the cognitive “black box” of active human–AI interaction. By analyzing data from a large qualitative sample (n = 55), this inquiry extends classical Cognitive Offloading Theory into high-order semantic and logical outsourcing, providing an empirical framework for process-focused pedagogy in smart learning environments.
PurposeThis meta-analysis aims to comprehensively review the impact of Generative Artificial Intelligence (Gen-AI) on college students’ critical thinking (CT) by quantitatively integrating the results of relevant empirical studies to obtain the overall effect. Design/methodology/approachThis meta-analysis synthesized data from 39 empirical studies published between 2023 and 2025. Effect sizes were calculated using random-effects models, and moderator analyses were conducted to examine potential influencing factors, including Gen-AI literacy level, disciplines, knowledge types, pedagogical approaches, user roles, Gen-AI interface types, Gen-AI roles, and Gen-AI task types. FindingsThe results indicated that Gen-AI had a moderately positive effect on CT (g = 0.591). Further analysis identified five significant moderating variables: disciplines, knowledge types, pedagogical approaches, Gen-AI roles and Gen-AI task types. Specifically, Gen-AI has the greatest positive impact on college students’ CT in STEM, procedural knowledge, inquiry-based learning, as a peer, and in the context of performing reflective and metacognitive tasks. These results suggest that within the overall contribution range of Gen-AI to college students’ CT, in some cases they may be more effective. Originality/valuePrevious research reviews, when exploring students’ higher-order thinking, did not make a clear distinction among the different types of thinking within them. Therefore, it is necessary to separate CT from broad learning outcomes or higher-order thinking and analyze its relationship with Gen-AI separately.
