AI Literacy as Both Bridge and Buffer: Unraveling Its Dual Role Between Research Stressors and Teaching Excellence

The persistent challenge of balancing research demands with teaching excellence in higher education is being reshaped by the advent of artificial intelligence, according to a new study published in Frontiers in Psychology. The research reveals that faculty’s proficiency with AI, termed "AI literacy," acts as a crucial tool, not only bridging the gap between academic pressures and pedagogical quality but also buffering the negative impacts of these pressures. This finding offers a potential pathway toward a more synergistic and sustainable model for higher education institutions grappling with the dual mission of advancing knowledge and educating future generations.

Research Findings Highlight AI’s Transformative Potential in Academia

A comprehensive survey of 253 faculty members from Chinese universities, conducted by researchers from Fuyang Normal University and Suzhou University of Science and Technology, investigated the intricate relationship between research stressors, teaching excellence, and the mediating role of artificial intelligence literacy. The study’s core findings underscore a nuanced understanding of how different types of research pressure affect teaching quality and how AI literacy can reconfigure these dynamics.

The research differentiated between "challenge stressors," which stem from demanding but growth-facilitating research tasks, and "hindrance stressors," which arise from job conditions that impede progress. Challenge stressors were found to have a significant positive impact on "Golden Course Quality" (GCQ), a benchmark for teaching excellence in China emphasizing high-order thinking, innovation, and challenge. This suggests that engaging with difficult research problems can directly enhance faculty’s pedagogical approaches, leading to more robust and innovative teaching.

Conversely, hindrance stressors, such as job insecurity or inadequate resources, did not directly impact GCQ. While these stressors did intensify the perceived conflict between teaching and research time, this conflict did not translate into a significant decline in teaching quality. This unexpected result challenges traditional views that resource depletion from research pressure inevitably harms teaching.

The study’s most significant contribution lies in its exploration of AI literacy’s "dual mechanism." Faculty with higher AI literacy were more adept at leveraging AI tools to enhance their teaching. AI literacy was found to mediate the positive effects of challenge stressors on teaching excellence, enabling faculty to translate research advancements into innovative course designs and challenging learning experiences. More critically, AI literacy acted as a buffer against the negative consequences of teaching-research time conflict. When faculty faced time constraints, those with stronger AI literacy could utilize AI to streamline tasks, personalize learning, and maintain high teaching standards, effectively decoupling time pressure from pedagogical output.

Contextualizing the Challenge: The "Research-First" Paradigm and Sustainable Development

Higher education globally is increasingly committed to achieving the United Nations Sustainable Development Goals (SDGs), with improving teaching quality recognized as a fundamental pillar for progress. However, many academic systems remain entrenched in a "research-first" paradigm. This prioritization, driven by performance evaluation metrics and university rankings, often leads to a situation where teaching investment is sidelined in favor of research output. This creates an unsustainable trend of "sacrificing teaching for research," impacting the quality of education provided.

In China, this challenge is particularly pertinent as the nation strives for "connotative development" in higher education. The Ministry of Education’s "Golden Course Construction Plan," launched in 2019, aims to elevate undergraduate education by fostering high-quality courses characterized by high-order thinking, innovation, and challenge—collectively known as "golden course quality" (GCQ). This initiative underscores the growing recognition of teaching innovation as a vital component of high-quality development, equal in importance to research.

Despite these policy shifts, the inherent tension between research and teaching persists. The widespread adoption of tenure-track systems has intensified research pressure, particularly on early-career faculty. This often leads to compromised instructional design, as the deep pedagogical preparation required for GCQ is overshadowed by the imperative to "publish or perish." Traditional theoretical frameworks, such as Conservation of Resources (COR) Theory, posit that this pressure leads to a depletion of faculty’s time and energy, ultimately diminishing teaching quality.

The Rise of Generative AI and Dynamic Capabilities

The emergence of generative artificial intelligence (GenAI) presents a paradigm shift, challenging the assumption of absolute resource scarcity that underpins COR Theory. Studies suggest that GenAI can automate core teaching tasks, significantly reducing preparation time and freeing up faculty to focus on more impactful pedagogical activities. Moreover, GenAI’s potential to provide personalized feedback and generate pedagogical insights can empower faculty to better balance their research and teaching responsibilities.

The ability of faculty to harness this potential hinges on their "AI literacy," conceptualized in this study as a dynamic capability. Drawing on Dynamic Capabilities Theory, AI literacy is understood as the ability to sense AI-enabled opportunities, seize them through technical proficiency, and reconfigure teaching resources to improve instructional practices. This dynamic capability framework suggests that technological empowerment can alter the traditional pathways through which research pressure affects teaching.

Methodology and Key Findings

The study employed partial least squares structural equation modeling (PLS-SEM) to analyze survey data from 253 faculty members. The model examined the direct and indirect effects of challenge and hindrance research stressors on GCQ, with teaching-research time conflict and AI literacy as mediating and moderating variables, respectively.

Key findings include:

• Challenge Stressors Enhance Teaching Excellence: Challenge research stressors directly and positively influenced GCQ. They also indirectly boosted GCQ by enhancing faculty AI literacy. This supports the idea that demanding research tasks, when perceived as developmental opportunities, can foster better teaching.
• Hindrance Stressors’ Indirect Impact: Hindrance stressors amplified teaching-research time conflict but did not show a significant direct negative effect on GCQ. This suggests that while these stressors create practical challenges, their impact on teaching quality is not as straightforward as previously assumed.
• AI Literacy as a Dual Enabler: AI literacy emerged as a critical factor. It mediated the positive pathway from challenge stressors to GCQ. Crucially, it buffered the negative impact of teaching-research time conflict on GCQ. Faculty with higher AI literacy were better able to manage time constraints and maintain teaching quality.
• Teaching-Research Time Conflict’s Weakened Link: The study found that teaching-research time conflict, while exacerbated by both types of stressors, did not significantly mediate the relationship between research stressors and GCQ. This indicates that the traditional resource depletion model might be less potent in an AI-augmented academic environment.

Theoretical and Practical Implications

The research offers significant theoretical advancements by extending COR Theory and Dynamic Capabilities Theory. It demonstrates that research pressure can be a catalyst for teaching improvement when it is challenge-oriented, and that AI literacy can disrupt the linear relationship between resource loss and performance outcomes. By reframing AI literacy as a dynamic capability rather than a mere technical skill, the study provides a new lens for understanding how faculty can adapt and thrive in the evolving academic landscape.

Practically, the findings offer actionable insights for higher education institutions. Administrators are encouraged to:

• Differentiate Stressor Management: Implement strategies that recognize the distinct effects of challenge and hindrance stressors. For challenge stressors, foster mechanisms that translate research insights into teaching innovations and incorporate these into performance evaluations. For hindrance stressors, focus on addressing root causes like resource scarcity and career uncertainty through targeted support and policy adjustments.
• Prioritize AI Literacy Development: Integrate AI literacy training into institutional strategies. This should be a tiered approach, equipping faculty with the skills to leverage GenAI for teaching-research integration and to mitigate time conflicts. This includes developing ethical guidelines and fostering critical evaluation of AI outputs.
• Foster a Synergistic Ecosystem: Move beyond a zero-sum view of research and teaching. Institutions should actively create an environment where these two missions can reinforce each other, emphasizing the development of dynamic capabilities like AI literacy to support faculty in navigating competing demands.
• Adapt Evaluation Frameworks: Consider incorporating "research integration intensity" into golden course evaluation systems, encouraging faculty to explicitly demonstrate how their research informs their teaching and enhances high-order thinking, innovation, and challenge.

Limitations and Future Directions

Despite its contributions, the study acknowledges several limitations. The sample, while diverse, was predominantly drawn from standard undergraduate colleges in China, potentially limiting generalizability to elite research institutions or cross-cultural contexts. Future research could benefit from multi-source evaluations (e.g., student ratings, peer reviews) to triangulate self-reported teaching quality and mitigate potential common-method bias. Furthermore, the use of two-item scales for GCQ dimensions, while ensuring content validity, might not fully capture the construct’s breadth. Future studies should develop comprehensive GCQ scales and employ longitudinal designs to track changes over time and monitor the evolving impact of GenAI. Exploring disciplinary differences in AI adoption and the role of organizational support and leadership in fostering AI literacy are also crucial avenues for future inquiry.

As higher education continues to navigate the complexities of the digital age, the insights from this research offer a promising roadmap for cultivating a more dynamic, resilient, and effective academic environment, ultimately contributing to the broader goals of sustainable development.