Objective This study aimed to classify profiles of elementary school students based on their levels of social-emotional competence and to examine the associations of demographic variables, including gender, grade, family structure on these latent profiles. Methods A stratified sampling method was employed to recruit 11,323 students from grades 3–6 across 42 schools in eastern, central, and western regions of China. Data were analyzed using Latent Profile Analysis (LPA) and Logistic regression. The four-profile solution was selected based on statistical indices (BIC = 1130587, Entropy = 0.912) and its interpretability. Results (1) elementary school students’ social-emotional competence could be classified into four categories: high level group (25.8%), medium to high level group (36.2%), low to medium level group (31.3%), and low level group (6.7%). (2) With the low level group as the reference category: compared with boys, girls were more likely to be classified into the high level group (OR = 15.701, P < 0.001); relative to Grade 3 students, Grade 4 students were more likely to be classified into the high level group (OR = 17.293, P < 0.001) and into the low to medium level group (OR = 4.061, P < 0.001); non-only children were more likely than only children to be classified into the high level group (OR = 15.679, P < 0.001) and into the low to medium level group (OR = 4.074, P < 0.001); students attending urban schools were more likely than rural students to be classified into the high level group (OR = 15.576, P < 0.001) and into the medium to high level group (OR = 4.084, P < 0.001); and non-left-behind children were more likely than left-behind children to be classified into the high level group (OR = 15.671, P < 0.001) and into the medium to high level group (OR = 4.094, P < 0.001). Conclusion Elementary school students exhibit distinct categorization patterns in social-emotional competence. These findings suggest that interventions should be tailored to address these profiles to enhance students’ social-emotional competence. It should be noted that these odds ratios are conditional on the low level group as the reference category. The low level group itself accounted for only 6.7% of the total sample, and within this already small category, the representation of certain demographic characteristics (e.g., girls, urban students) was extremely low. These two factors combined contribute to the large magnitudes of the reported odds ratios. Therefore, these values represent relative advantages compared to the low level group rather than overall population odds ratios, and should be interpreted cautiously.
Gao et al. (Wed,) studied this question.