學齡前兒童執行功能與奮力控制氣質之雙領域潛在成長軌跡探究

Exploring Dual-Domain Latent Growth Trajectories of Executive Function and Effortful Control Temperament in Preschool Children

本研究探討臺灣學齡前兒童兩項核心能力─執行功能與奮力控制氣質─的發展軌跡及其相互關係。研究資料取自「臺灣幼兒發展調查資料庫」，共納入1，694名36至72月齡兒童的四波縱貫資料，並運用雙領域潛在成長模型進行分析。研究結果顯示，執行功能呈現穩定的線性成長，而奮力控制氣質則在學齡前期迅速成長，隨後成長速率逐漸趨緩。兩項能力從36月齡起即顯現出顯著的個體間差異，顯示兒童在自我調節發展上具有高度變異性。此外，研究發現執行功能與奮力控制氣質的起始能力與後續成長率之間皆呈現負相關，亦即初始能力較低的兒童在發展上反而有較大的成長幅度。更進一步地，兩者的起始點與成長率之間皆存在顯著正相關，顯示出執行功能與奮力控制氣質具有發展同步性與交互促進的潛能。此一發現補足了臺灣在早期兒童自我調節雙要素共同發展方面的縱貫研究缺口。在實務意涵上，本研究指出，早期介入應同時涵蓋執行功能與氣質等二個能力，以促進學齡前兒童的身心發展。

Preschool is a pivotal developmental period during which children's self-regulatory capacities expand rapidly and lay the groundwork for later cognitive, social, and emotional adaptation. Two constructs have been consistently highlighted as core components of early self-regulation: Executive function (EF) and effortful control (EC). Although EF and EC originate from different theoretical traditions (neurocognitive vs. temperament), accumulating evidence suggests that they are distinguishable yet strongly related and jointly predictive of children's adaptive functioning. In early childhood, both EF and EC have been linked to academic readiness, emotional and behavioral regulation, peer competence, and long-term adjustment. Despite this importance, there remains a need for longitudinal evidence clarifying how EF and EC develop together over time, whether their growth patterns are parallel or distinct, and how individual differences in initial levels relate to subsequent developmental change-especially in Taiwanese children using nationally representative longitudinal data. Guided by developmental and transactional perspectives on self-regulation, the present study investigated (a) the developmental trajectories of preschool children's EF and EC from 36 to 72 months of age, (b) inter-individual differences in initial status and growth rates for both domains, and (c) the extent to which EF and EC exhibit synchronized development or compensatory dynamics across time. To address these aims, we analyzed four-wave longitudinal data drawn from the Kids in Taiwan: National Longitudinal Study of Child Development & Care (KIT), a large-scale cohort study designed to capture Taiwanese children's development across early childhood. The analytic sample included 1,694 children followed at 36, 48, 60, and 72 months. This dataset is particularly valuable because its sampling design, measurement development, and quality control procedures support robust developmental inference and offer an empirical basis for informing early intervention and educational practice in Taiwan. Methodologically, we employed a dual-domain latent growth modeling (LGM) framework to examine EF and EC simultaneously. This approach allows researchers to estimate mean developmental change (fixed effects), individual differences in initial levels and growth (random effects), and cross-domain associations between growth factors (e.g., intercept-intercept, slope-slope, and intercept-slope relations) within a single integrated model. Prior to growth modeling, we evaluated longitudinal measurement invariance to ensure that EF and EC were assessed consistently across the four time points. Following commonly used criteria for invariance evaluation, changes in comparative fit index (ΔCFI) and root mean square error of approximation (ΔRMSEA) supported strong (scalar) invariance across waves, indicating that observed changes could be interpreted as developmental change rather than measurement artifacts. This step is essential in longitudinal research because meaningful interpretation of growth parameters presupposes that the constructs retain the same meaning over time. Results from the dual-domain LGM revealed clear yet distinct developmental patterns for EF and EC across the preschool years. On average, EF demonstrated a stable, approximately linear increase from 36 to 72 months, consistent with the view that rapid maturation of cognitive control processes continues throughout early childhood. In contrast, EC showed a pattern characterized by relatively rapid growth earlier in the preschool period followed by a gradual deceleration, suggesting a nonlinear developmental process. This pattern aligns with temperament theory emphasizing that EC is not fully established in infancy but emerges and consolidates during toddlerhood and preschool, becoming increasingly stable as children approach school age. Taken together, these findings indicate that while both EF and EC improve across early childhood, they may do so through somewhat different growth functions-one more consistently incremental (EF) and the other more stage-sensitive (EC). Crucially, substantial inter-individual variability was observed in both domains. Variances for EF and EC intercepts were significant, indicating meaningful differences among children in their initial levels as early as 36 months. Variances for growth factors were also significant, showing that children differed not only in where they started but also in how quickly they developed across the preschool years. This heterogeneity underscores a key implication for early childhood policy and practice: Developmental pathways of self-regulation are not uniform, and children may require differentiated supports depending on their starting points and growth potential. Such results are consistent with developmental systems perspectives emphasizing that children's competencies emerge from ongoing interactions between person characteristics and contextual experiences. Within-domain relations further showed a pronounced 'catch-up' pattern. The correlation between EF intercept and EF slope was significantly negative, and the correlation between EC intercept and EC slope was also significantly negative. These coefficients indicate that children who started lower at 36 months tended to exhibit steeper subsequent growth, whereas those who started higher tended to show slower growth-suggesting diminishing returns or ceiling effects as competencies strengthen. This 'low-start, high-growth' phenomenon is developmentally meaningful: It implies that early disadvantage in EF or EC does not necessarily lock children into persistently low trajectories, and that many children may demonstrate accelerated gains when developmental demands and learning opportunities increase during preschool. Cross-domain associations demonstrated both synchrony and dynamic interplay between EF and EC. First, the intercept- intercept correlation was significantly positive, indicating that children with higher EF at 36 months also tended to show higher EC at the same age. This co-occurrence supports the conceptualization of EF and EC as interrelated aspects of early self-regulation, despite their distinct theoretical origins. Second, the slope-slope correlation was also significantly positive, suggesting coordinated development: Children who improved more rapidly in EF also tended to show faster growth in EC across 36-72 months. This synchronized growth pattern is consistent with claims that neurocognitive control and temperamental self-regulation may mutually reinforce each other during early development, potentially through shared reliance on attentional control and inhibitory processes and through overlapping environmental supports (e.g., structured routines, responsive guidance, opportunities for rule-based play). In addition to synchronization, evidence also pointed to compensatory cross-domain dynamics involving initial status and growth. Specifically, cross-construct intercept-slope relations (e.g., EF intercept with EC slope; EC intercept with EF slope) were significantly negative, implying that lower initial competence in one domain was associated with steeper growth in the other. This pattern suggests a form of functional compensation within the broader self-regulation system, whereby children may rely more heavily on alternative regulatory processes when a particular component is initially weaker. Such compensatory dynamics align with perspectives emphasizing that self-regulation is supported by multiple interacting subsystems and that developmental cascades can occur across domains over time. In practical terms, this finding implies that interventions targeting one domain (e.g., EF) may yield broader benefits by mobilizing or strengthening complementary regulatory capacities (e.g., EC), and vice versa. Overall, this study contributes to the literature in three primary ways. First, using large-scale Taiwanese longitudinal data, it provides local empirical evidence that both EF and EC show meaningful developmental change from 36 to 72 months, while also highlighting distinct growth patterns (linear vs. non-linear). Second, it documents substantial heterogeneity in initial levels and developmental rates, reinforcing the importance of individualized perspectives in early childhood education and intervention. Third, by adopting a dual-domain growth framework, it clarifies how EF and EC co-develop-demonstrating both synchronized growth (positive intercept-intercept and slope-slope associations) and compensatory linkages (negative intercept- slope relations), thereby extending understanding beyond static correlations. These findings yield several implications for educational practice and early intervention. Because individual differences are evident by 36 months and growth is most dynamic during the preschool years, early identification and support are crucial. Interventions should not focus exclusively on either EF or temperament-based regulation; rather, integrated approaches that concurrently strengthen cognitive control (e.g., working memory and inhibitory games, rule-switching activities) and temperamental self-regulation (e.g., attention focusing, delay of gratification, emotionally supportive scaffolding) may be most effective. Moreover, the observed catch-up and compensatory patterns indicate that children with lower early competence may have substantial developmental plasticity, making preschool a particularly promising window for prevention and promotion efforts. Several limitations should be noted. Although longitudinal measurement invariance was supported, EF and EC were assessed through questionnaire-based indicators, which may be influenced by rater perceptions and contextual bias. Future studies could incorporate multi-method assessments (e.g., behavioral tasks, observational measures, and physiological indicators) to strengthen validity. In addition, while latent growth modeling captures developmental change, causal inference remains limited; future research could examine potential mechanisms (e.g., parenting, classroom quality, socioeconomic contexts) that shape individual differences in trajectories and the coupling between EF and EC. Extending the time span into the early school years would also help clarify whether the synchronized and compensatory dynamics observed in preschool persist, strengthen, or reorganize as academic and social demands increase. In conclusion, using nationally based longitudinal data and a dual-domain latent growth approach, this study demonstrates that Taiwanese preschool children's EF and EC both develop substantially between 36 and 72 months, with clear interindividual variability and meaningful cross-domain coupling. The results underscore that early self-regulation is best understood as a coordinated system involving both cognitive and temperamental components, and they highlight the value of integrated early supports to promote children's adaptive development.