Abstract: Since China' s “Double Reduction” policy was implemented in 2021, educational focus has broadened from solely academic outcomes to include students' learning processes. However, few longitudinal studies have examined how this policy has altered parental educational involvement and how such changes subsequently influence students' learning engagement and academic performance. Parental involvement is a key determinant of children' s academic outcomes. The “Double Reduction” policy, while reducing student burdens, also redefines parental roles, potentially altering the quality and dimensions of their involvement. Yet, prior research lacks longitudinal evidence characterizing the developmental trajectories of its three core subdimensions (i.e., emotional, intellectual, and behavioral involvement) across the policy transition period. This longitudinal study addresses this gap by investigating how baseline levels and change rates of these subdimensionspredict students' learning engagement and academic achievement two years later.
We recruited a cluster sample of 323 primary school students from two schools in Shandong Province, China. Data were collected at three waves: baseline (June 2021, Wave 1), 1.5 years later (Wave 2), and 2 years later (Wave 3). Participants' mean age at baseline was 8.89 years (SD = 0.94, range = 7 to 11 years old; 47.82% boys). Parental involvement was measured using a validated Parental Involvement Scale across all three waves. Students' learning engagement was assessed via self-report at Waves 1 and 3 using a Learning Engagement Scale. Academic achievement (final grades) was obtained from school records at Wave 3. Demographic covariates were collected at baseline. We applied latent growth modeling in Mplus 8.3 to (a) estimate unconditional linear growth trajectories for emotional, intellectual, and behavioral involvement across the three waves; (b) estimate individual differences in intercepts (initial levels) and slopes (change rates) and their interrelations; and (c) test conditional latent growth models, including interactions between intercepts and slopes, to predict Wave 3 learning engagement and academic achievement, controlling for baseline age and learning engagement.
Several findings were obtained. First, unconditional latent models showed that parental involvement's sub-dimensions (emotional, intellectual, and behavior involvement) exhibited a linear decreasing trend before and after the implementation of the “double reduction”. Second, the individual differences in parental involvement among primary school parents demonstrated a stable developmental pattern over time. Third, after controlling for age and learning engagement at Wave 1, conditional latent growth models showed that emotional and behavior involvement had cumulative effects on students' learning engagement under the background of “double reduction.” In terms of academic performance, the change rates of emotional involvement significantly predicted academic performance, while intellectual involvement showed no significant effects on either learning engagement or academic performance. Supported by the amplification model, students with a low initial level of parental behavioral involvement experienced a more rapid decline in academic performance when the decline rate of behavior involvement accelerated. In contrast, students with high initial levels of behavioral involvement experienced slower declines in academic performance, even when they experienced faster rates of decline.
The present study makes both theoretical and practical contributions. First, this study provided a comprehensive overview of the developmental trajectory of parental involvement before and after the “double reduction” policy among Chinese primary school children by examining initial levels, change rates, and their interrelationships. Second, these results underscore the differential effects of the initial level and growth rates of different dimensions of parental involvement on learning engagement and on academic achievement. Third, this study emphasizes the importance of prioritizing both the initial level and change rate of behavioral involvement in fostering children's learning and the necessity of high-quality parental involvement for academic development. Practically, it highlights the critical importance of sustaining high-quality behavioral involvement and provides evidence for tailoring family support strategies and school-home collaboration in the post-policy context.

Key words: academic performance, “double reduction” policy, latent growth modeling, learning engagement, parental involvement