The purpose of this study is to identify transition patterns in self-management competence during the middle-to-high school transition and to explore key predictors of membership in transition groups exhibiting change. Previous research has primarily ...
The purpose of this study is to identify transition patterns in self-management competence during the middle-to-high school transition and to explore key predictors of membership in transition groups exhibiting change. Previous research has primarily relied on cross-sectional data and samples of college students, which limits the ability to capture how self-management competence among middle and high school students changes differently across individuals depending on developmental stage and contextual factors. To address this limitation, this study focused on the school-level transition period and employed latent transition analysis (LTA) to identify transition patterns in self-management competence. To this end, LASSO (least absolute shrinkage and selection operator)-based variable selection was employed to identify predictors of membership in transition groups exhibiting change, and logistic regression analyses were subsequently conducted to examine the statistically significant effects of the selected predictors on changes in self-management competence.
The specific research questions are as follows:
1. What are the transition patterns of self-management competence profiles during the middle-to-high school transition?
1-1. What are the profiles of self-management competence among students in the second grade of middle school?
1-2. What are the profiles of self-management competence among students in the first grade of high school?
1-3. What are the transition types of self-management competence profiles between the second grade of middle school and the first grade of high school?
2. What are the key predictors of transitions in self-management competence profiles during the middle-to-high school transition?
This study utilized data from the Korean Educational Longitudinal Study2013 (KELS2013), specifically the fourth wave (the second grade of middle school) and the sixth wave (the first grade of high school), to examine transition patterns in self-management competence profiles across the middle-to-high school transition. Using profiles identified in the second grade of middle school as a baseline, the study further explored predictors distinguishing students who maintained their self-management competence from those who experienced changes. Predictors were constructed to include variables at the student, family, and school levels. The final analytic sample consisted of 5,239 students who, by the sixth wave, were attending general high schools, autonomous high schools, or specialized high schools.
The major findings of the study are summarized as follows.
First, latent profile analysis identified three distinct self-management competence profiles at both the second grade of middle school and the first grade of high school: low, medium, and high levels of self-management competence. These profiles differed primarily in quantitative levels rather than qualitative patterns across subdomains. Results from the latent transition analysis indicated that during the transition from middle to high school, the majority of students remained in the same profile. However, approximately 20–30% of students exhibited upward or downward transitions in self-management competence profiles during the school transition period, with most changes occurring between neighboring levels.
Second, LASSO-based variable selection revealed that, across all three groups defined by initial levels of self-management competence in the second grade of middle school (low, medium, and high), variables related to learning motivation and self-regulation strategies were consistently selected, indicating that self-regulated learning characteristics constitute core predictors of changes in self-management competence. However, the structure of predictors differed depending on the initial level of self-management competence. Logistic regression analyses showed that, in the low-level group, the strengthening of individual internal resources predicted upward transitions; in the medium-level group, individual, and school factors jointly contributed to changes; and in the high-level group, declines in motivation and self-regulation emerged as key predictors of downward transitions.
This study empirically demonstrated that patterns of transition in self-management competence and their predictors during the middle-to-high school transition differ according to students’ initial levels, suggesting that the development of self-management competence should be understood not as a single average trajectory but as a set of group- and level-specific developmental processes. The findings further indicate that self-management competence is more likely to be maintained and gradually adjusted than to change abruptly during the school transition period, highlighting the importance of focusing transitional support on the sustained use and incremental refinement of existing competencies rather than short-term interventions. Methodologically, this study explores the integration of machine learning–based variable selection and traditional regression analysis to examine predictors of change in self-management competence from an interpretability perspective, representing an attempt to bridge prediction-oriented and interpretation-oriented approaches.
Future research should examine heterogeneous patterns of change in self-management competence during the transition between school levels by distinguishing multiple transition pathways defined by the combination of initial and final levels. From a methodological perspective, the use of multi-year longitudinal data would enable a more refined analysis of developmental processes and individual differences in trajectories before and after the transition period. In this regard, self-management competence represents a complex construct integrating cognitive, emotional, and behavioral components and is shaped through interactions between individual characteristics and environmental contexts. Accordingly, predictions of transition status have a limited scope of interpretation and should be interpreted with caution.