Investigating changes in university students’ academic engagement and its relationships with motivation and performance: a quasi-longitudinal study based on latent change score analysis during the COVID-19 pandemic

Academic engagement (AE) has been widely recognized as a key determinant to address key challenges in the academic context as the student persistence and success at university. However, previous studies have mostly measured AE using cross-sectional data, so its development over time is still unexplored. To address this research gap, we involved 209 first‑year STEM undergraduates (54% female) from April–May 2019 (t₀) to October–November 2021 (t₁) and investigated: (i) how AE changed from first to third year; (ii) how these changes in AE were affected by initial motivation; and (iii) how they affected students’ academic performance. Using the six dimensions of SInAPSi Academic Engagement Scale (SAES) alongside the Relative Autonomy Index (RAI) for motivation and total ECTS credits for performance, latent change score (LCS) models were performed to answer the research questions. In general, we found that a lower initial score was associated with greater changes, thus supporting AE as a dynamic process rather than a trait construct. Results indicated a significant increase in all dimensions except for engagement with university colleagues dimension. Moreover, higher autonomous regulations led to higher initial engagement, which remained stable over time. Finally, higher levels of engagement with colleagues and capability to persist in the first year were associated with better performance in the third year. Findings necessarily reflected effects of instruction disruption during COVID-19 pandemics. The study contributes to the debate about the role of engagement as a multidimensional dynamic construct influenced by social connections, that in turn affect academic achievement.