The study is grounded in conservation of resources theory, an influential framework in psychology developed by Stevan Hobfoll. The theory proposes that people strive to obtain, retain, and protect the resources they value, and that psychological well-being depends on having enough resources in reserve to meet demands. In an educational setting, academic buoyancy functions as a core personal learning resource: it helps students withstand the everyday stresses of coursework without depleting their motivational reserves. Learning engagement, by contrast, is the visible output of a well-resourced learner, encompassing the energy, enthusiasm, and absorption students bring to their studies. According to the theory’s logic, students who possess a rich resource like buoyancy should be better positioned to invest in engagement, and the gains in one domain should reinforce gains in the other, creating an upward spiral of development.

To test this idea rigorously, Zhang designed a repeated-measures longitudinal study rather than relying on the one-shot questionnaires that dominate much of the second-language acquisition literature. Participants were college students learning English as a foreign language in China, a context where learners face constant daily challenges: high-stakes examinations, dense vocabulary demands, and limited opportunities to use English outside the classroom. Over the sixteen-week semester, the students completed the same questionnaire batteries on multiple occasions, allowing the researcher to chart not just whether buoyancy and engagement were high or low, but how each construct changed week by week across an entire term of instruction.

The analytical centerpiece of the study is latent growth curve modeling, or LGCM, a structural equation modeling technique that treats individual change over time as a latent variable in its own right. Instead of simply comparing average scores between the beginning and end of the semester, LGCM estimates two key parameters for each person: an intercept, which captures that individual’s starting level when the study began, and a slope, which captures the rate at which that individual’s scores rose or fell across the measurement occasions. Because the intercepts and slopes are modeled as latent variables, the approach separates genuine developmental change from measurement error and makes it possible to test how one construct’s trajectory relates to another’s with remarkable statistical precision.

The first major finding concerns the shape of the trajectories themselves. Across the semester, both academic buoyancy and learning engagement exhibited overall upward trends, meaning that, on average, students became more resilient and more engaged as the term progressed. Crucially, however, the averages concealed substantial individual variability. Some students started the semester with high buoyancy and grew even stronger, while others began low and stayed there; the same was true for engagement. This heterogeneity matters because it tells educators that a single classroom-level snapshot of motivation tells only a fraction of the story. The students sitting in the same lesson may be on radically different developmental paths, and interventions that help one learner may do little for another.

The second set of findings illuminates how the starting point of each construct shaped its own development. For academic buoyancy, the results showed a clear pattern: students whose latent intercept was higher at the beginning of the study experienced greater subsequent growth in the construct over the semester. In other words, buoyancy appeared to beget buoyancy, a Matthew effect in which students who already possessed this psychological resource accumulated more of it over time. Strikingly, the same pattern did not hold for learning engagement. The initial state of engagement showed no significant association with its own developmental trajectory, suggesting that starting the semester highly engaged did not guarantee that engagement would keep climbing. Engagement, on this evidence, is not self-sustaining in the way buoyancy appears to be; it needs an external engine.

That engine, the third and arguably most consequential finding suggests, is academic buoyancy. The latent intercept of buoyancy was positively associated with both the intercept and the growth rate of learning engagement. Students who began the semester more buoyant were more engaged from the outset, and, just as importantly, they showed steeper growth in engagement as the weeks passed. Furthermore, growth in academic buoyancy was aligned with parallel growth in engagement, so that as individual students became better at bouncing back from daily setbacks, their classroom engagement climbed in tandem. The data are correlational and cannot prove causation, but the temporal ordering and the parallel trajectories are exactly the pattern that conservation of resources theory predicts when a personal resource fuels sustained motivational investment.

For the second-language teaching community, the practical implications are considerable. Much research and policy attention has focused on engagement as the thing to be engineered: better tasks, more interaction, richer input. This study suggests that engagement may be downstream of a more fundamental capacity, and that programs which explicitly build students’ ability to handle everyday academic adversity, through feedback practices, normalizing mistakes, teaching coping strategies, and structuring challenges at manageable difficulty, could pay dividends that compound across a semester. The finding that initial buoyancy predicted its own growth adds urgency: learners who lack this resource early may fall further behind motivationally, and early identification could allow educators to intervene before the gap widens. At the same time, the malleability of both constructs, evident in their upward trajectories over a single term, is an optimistic signal that these qualities can be developed rather than being fixed traits.

Methodologically, the study also makes a case for a broader shift in how researchers study motivation in second language acquisition. Cross-sectional designs, which measure constructs at a single point in time, cannot distinguish a stable disposition from a developmental trend, and they cannot reveal whether two variables that correlate at one moment also grow together. Latent growth curve modeling, by contrast, treats development as the object of study. Zhang’s demonstration that the technique can uncover asymmetric dynamics, with buoyancy’s initial level shaping its own growth while engagement’s initial level does not, shows the kind of insight that only longitudinal modeling can deliver. As the field moves toward viewing language learning as a dynamic system in which psychological variables fluctuate and interact over time, tools of this kind are likely to become standard equipment.

The research leaves open questions that future work should address, including whether the same buoyancy-to-engagement pathway operates in other languages, age groups, and educational cultures, and whether experimental interventions that raise buoyancy produce downstream gains in engagement. Yet the core message is already clear and striking. The everyday resilience of language learners, their quiet ability to absorb a bad grade or an awkward classroom moment and carry on, is not merely a nice-to-have personality feature. It is a measurable, growing psychological resource that tracks, predicts, and travels alongside the very engagement that language learning ultimately depends on. For students grinding through a difficult semester of English study, the capacity to bounce back may be the single most valuable skill they never knew they were practicing.

Subject of Research: The longitudinal relationship between academic buoyancy and learning engagement in English-as-a-foreign-language learners

Article Title: Academic buoyancy and engagement in EFL learning: a latent growth curve modeling analysis of their dynamic relationship

Article References: Zhang, T. (2026). Academic buoyancy and engagement in EFL learning: a latent growth curve modeling analysis of their dynamic relationship. Frontiers in Psychology, 17, Article 1944779. https://doi.org/10.3389/fpsyg.2026.1944779

Image Credits: AI Generated

DOI: 10.3389/fpsyg.2026.1944779

Keywords: academic buoyancy, learning engagement, EFL learners, latent growth curve modeling, conservation of resources theory, longitudinal study, second language acquisition, educational psychology, student motivation, China, Academic, buoyancy