Engaged Students Feel Better: Five-Month Study Tracks Well-Being in Austrian Fifth-Graders

When children start secondary school, they face new classrooms, new teachers and new social worlds — and researchers have long suspected that how engaged those children feel at school shapes how happy they are, and vice versa. A new longitudinal study from Austria now offers some of the clearest evidence yet on which direction the arrow points. Following 1,762 fifth-grade students across two points in a single school year, Sebastian Redl of the Linz School of Education at Johannes Kepler University Linz found that student engagement predicted later well-being more consistently than the reverse. The study, published in Frontiers in Education, provides partial support for a celebrated psychological theory of positive emotions — and a nuanced picture of how the two constructs feed into each other during one of the most turbulent transitions in a child’s school career.

The research takes its cue from positive education, a movement rooted in positive psychology that argues well-being is not merely a pleasant by-product of schooling but a central ingredient of academic success. Previous studies have repeatedly linked well-being to achievement, attendance, self-regulation and engagement, and rising mental health problems among adolescents in Western countries have sharpened the urgency of understanding these connections. Yet most of that evidence rests on cross-sectional snapshots, which cannot establish whether happy students become engaged students, engaged students become happy students, or both. Longitudinal designs are the exception in this literature, and the few bidirectional studies that exist have mostly been conducted with Asian samples, leaving European data scarce.

The study also tackles a subtler problem: lumping constructs together. Subjective well-being, the hedonistic variety of well-being used here, is defined as people’s beliefs and feelings about whether they are leading a desirable and rewarding life, and it comprises three distinct dimensions — positive affect, negative affect and life satisfaction. Student engagement, meanwhile, is typically modelled as a meta-construct with behavioural, emotional and cognitive facets: following class rules and contributing to discussions, feeling interest or boredom toward school, and investing effort strategically in learning. Earlier work that collapsed these dimensions into composite scores could not detect whether, say, positive affect drives engagement while life satisfaction does not. Some studies have indeed found divergent patterns, including one Asian high-school sample in which positive affect, unlike the other two well-being dimensions, showed no influence on engagement.

The theoretical engine of the study is the broaden-and-build theory and its extension, the upward spiral hypothesis. The theory holds that positive emotions such as joy, interest and contentment widen people’s thought–action repertoires — the urge to play, explore, push limits or integrate — and thereby build lasting physical, psychological, intellectual and social resources. Negative emotions such as anxiety, sadness and anger do the opposite, narrowing the repertoire to quick, threat-avoiding actions, though the theory is deliberately asymmetrical: negative emotions need not actively destroy resources, they may simply prevent their accumulation. Positive emotions can also shorten stress reactions caused by negative ones, the so-called undoing hypothesis. The upward spiral hypothesis then predicts a dynamic reciprocal loop, in which resources built by positive emotions make future positive emotions more likely.

To test these ideas, Redl surveyed fifth-graders from 40 schools in Upper Austria in January/February 2025 and again in June 2025, roughly five months apart. Fifth grade marks the transition from primary to secondary school in Austria, a period accompanied by new requirements, increased performance pressure and changed social dynamics, and a developmental stage when school motivation and engagement often decline. The 1,494 children who provided data at both waves had a mean age of 10.81 years; 42 percent were girls and 49 percent boys. Well-being was measured with the German student life satisfaction scale and the short form of the Children’s Positive-and-Negative Affect Schedule, while engagement was captured with 15 items from the multidimensional school engagement scale, translated and back-translated into German to ensure linguistic equivalence.

The statistical machinery was a cross-lagged panel model, which estimates each variable’s stability over time (autoregression) while simultaneously testing whether one construct at Time 1 predicts another at Time 2, controlling for all other relationships. Confirmatory factor analyses confirmed good fit for the well-being model at both time points, and longitudinal measurement invariance held for both constructs, meaning they carried the same meaning in January as in June. Notably, the engagement data revealed substantial overlap among the behavioural, emotional and cognitive dimensions: once a general engagement factor was added, the behavioural and cognitive specific factors retained virtually no unique variance and were dropped, leaving a bifactorial model of general engagement plus a specific emotional engagement factor, with a method factor for negatively worded items. Attrition analyses showed no significant influence of any construct on dropout, and full information maximum likelihood handled missing data, with cluster-robust standard errors accounting for the nesting of students in school classes.

The results delivered a clear asymmetry. Positive affect early in the school year significantly predicted later general engagement (β = .097, p = .034), though not the specific emotional engagement factor. Life satisfaction, by contrast, showed no significant effect on either engagement factor, and negative affect predicted neither — so the hypotheses that well-being drives engagement were only partially supported. In the reverse direction, the evidence was stronger: both the general engagement factor (β = .085, p = .009) and the specific emotional engagement factor (β = .057, p = .046) positively predicted later life satisfaction, and both also predicted later positive affect (β = .107, p = .001 for the general factor; β = .085, p = .009 for the specific factor), fully supporting the upward spiral predictions. Neither engagement factor, however, predicted later negative affect.

When Redl compared the opposing cross-lagged paths directly, using the Delta method to derive standard errors for the differences between fully standardised coefficients, no pair differed significantly. Descriptively, engagement was the somewhat stronger predictor of later well-being than the reverse, but the differences were small (all |Δβ| ≤ .10) and statistically insignificant. The study therefore cannot claim a decisive directional asymmetry — only that the reciprocal loop, where it exists, runs through positive affect and general engagement, and that negative affect appears largely detached from the cycle over this five-month window. Student engagement also proved more temporally stable than subjective well-being across the school year.

What do these patterns mean theoretically? The upward spiral hypothesis received support in a younger age group than previously studied: higher engagement, treated as a personal resource, fostered later positive emotions. The broaden-and-build theory was partially supported, since positive affect boosted general engagement but not emotional engagement beyond it. Redl suggests that positive affect may primarily promote general arousal, cognitive flexibility and perseverance that manifest as common variance across all engagement indicators, while the emotional engagement that remains specific may depend more on situational factors such as interest or a student’s current social integration in the class. The absence of negative-affect effects is consistent with other studies in which positive emotions shaped engagement but negative ones did not, and it may reflect the undoing hypothesis — positive emotions cancelling out negative influences — or simply the short time lag of the study.

The author is careful about limitations. Two measurement points cannot exclude alternative explanations, and at least three would be needed to model more complex dynamics; all data were self-reported, inviting self-report and monomethod bias; the sample is exclusively Austrian, so generalisation to other sociocultural settings is uncertain; and the Delta-method comparisons are approximations that future work should replicate with bootstrapping. Still, the practical implication is straightforward: because engagement and positive emotions reinforce each other, proactively promoting either — in a developmental window when mental health problems are rising — could act as a preventive measure reaching many children at once, nurturing both the joy and the investment that carry students through the difficult transition to secondary school.

Subject of Research: Longitudinal relationships between student engagement and subjective well-being in lower secondary education

Article Title: Analysing the link between student engagement and subjective well-being in lower secondary education: evidence from a cross-lagged panel study

Article References: Redl, S. (2026). Analysing the link between student engagement and subjective well-being in lower secondary education: evidence from a cross-lagged panel study. Frontiers in Education, 11, 1968024. https://doi.org/10.3389/feduc.2026.1968024

Image Credits: AI Generated

DOI: 10.3389/feduc.2026.1968024

Keywords: student engagement, subjective well-being, positive education, broaden-and-build theory, upward spiral hypothesis, cross-lagged panel model, early adolescence, positive affect, life satisfaction, negative affect, Austria, longitudinal study