Modern education increasingly leverages information technology to enhance learning. Three-dimensional (3D) learning resources, derived from the computer game industry, offer interactive and immersive learning environments. These resources, encompassing 3D videos, animations, virtual reality (VR), and augmented reality (AR), aim to improve spatial understanding and knowledge acquisition. While studies have shown the cognitive benefits of 3D learning resources, their impact on learning resilience—the ability to adapt positively to learning challenges—remains under-researched. This study addresses this gap by examining the effect of 3D learning resources on college students' learning resilience, focusing on the mediating roles of positive emotion (based on the broaden-and-build theory) and cognitive load (based on cognitive load theory). The research questions are: 1. Does the use of 3D learning resources impact students' learning resilience? 2. How do 3D learning resources affect students' learning resilience? The study aims to provide insights into enhancing teaching quality and promoting learning resilience among college students, addressing the issue of high dropout and failure rates. This research offers a novel theoretical model evaluating the influence of 3D learning resources on learning resilience through three distinct paths, leading to both expected and novel insights in the existing knowledge system.

The concept of resilience, initially from physics, evolved in psychology to describe the ability to adapt to adversity. Learning resilience, specifically, focuses on learners' ability to overcome learning challenges. Numerous factors influence learning resilience, including individual cognitive processes, family environment, educational institutions, and societal influences. Positive emotions are crucial, with optimistic emotions facilitating learning and negative emotions creating barriers. External factors such as family warmth, parental support, teacher encouragement, and classroom environment also play significant roles. While research exists on measuring and analyzing learning resilience, comprehensive studies on the impact mechanisms of 3D learning resources on learning resilience in higher education are lacking. This study fills this gap by empirically investigating these mechanisms.

The study employed a quantitative approach using a survey design. Data were collected from 963 college students in Shandong Province, China, who were using 3D digital resources in their classrooms. A seven-point Likert scale questionnaire measured eight variables: information quality, system quality, service quality (all aspects of 3D learning resources), positive emotion, cognitive load, perseverance, reflective and adaptive help-seeking, and negative affect and emotional response (all aspects of learning resilience). 3D learning resources and learning resilience were treated as second-order factors, each composed of three first-order factors. A pilot study was conducted to refine the questionnaire. Convenience sampling was used. Data analysis involved reliability and validity tests (KMO, factor analysis, Cronbach's α, composite reliability, AVE, discriminant validity), homogeneity tests (ANOVA), common method bias tests (Harman's single factor test), and structural equation modeling (SEM) using Lisrel 8.7 and Mplus 7.0 software (for mediation effect testing using bootstrapping).

The findings support five hypotheses. First, 3D learning resources positively influence learning resilience (β = 0.25, p < 0.001). Second, 3D learning resources positively influence positive emotion (β = 0.89, p < 0.001). Third, positive emotion positively influences learning resilience (β = 0.68, p < 0.001). Fourth, 3D learning resources negatively influence cognitive load (β = -0.15, p < 0.001). Fifth, cognitive load negatively influences learning resilience (β = -0.06, p < 0.001). Furthermore, both positive emotion and cognitive load act as significant mediators in the relationship between 3D learning resources and learning resilience. Mediation effect confidence intervals for positive emotion and cognitive load were [0.067, 0.121] and [0.042, 0.090], respectively, neither including zero. The correlations between 3D learning resources, cognitive load, and learning resilience were relatively weak, possibly due to the influence of other factors on cognitive load and learning resilience beyond 3D resources.

The findings demonstrate that 3D learning resources have a positive impact on college students' learning resilience, both directly and indirectly through the mediating effects of positive emotion and cognitive load. The positive effect of 3D resources aligns with existing research highlighting their engagement and interactivity. The mediating role of positive emotion supports the broaden-and-build theory, emphasizing the importance of positive emotional states in fostering resilience. The negative mediating effect of cognitive load underscores the need for well-designed 3D resources that minimize cognitive overload. The relatively weak correlations highlight the multifaceted nature of cognitive load and learning resilience, influenced by factors beyond 3D learning resources. This study's findings extend existing knowledge by providing empirical evidence for a comprehensive model explaining the relationship between 3D learning resources and learning resilience.

This study provides empirical support for the positive impact of 3D learning resources on learning resilience, mediated by positive emotion and cognitive load. The findings have implications for educational practice, suggesting that universities should invest in developing and implementing high-quality interactive 3D resources. Future research should explore longitudinal effects, include additional factors influencing resilience, and account for technological advancements. The study's contribution is the development of a novel theoretical model that integrates the direct and indirect effects of 3D learning resources, enhancing our understanding of learning resilience.

The study's limitations include its cross-sectional design, limiting the ability to establish temporal relationships between variables. The use of only a questionnaire may have limited the richness of the data. Further, the focus solely on 3D resources ignores other potential factors influencing learning resilience, and the study is limited to the current technological context, which might change with future technological advancements.