(WITHDRAWN) Similarities and Differences in the Measurement of Youth Social-Emotional Competence across Eastern and Western Countries

Background:

Social-emotional competence (SEC) is a multidimensional construct involving diverse skills and mindsets enabling healthy interactions with oneself and others. Despite increasing efforts to promote youth SEC worldwide, the current literature on SEC measurement is largely based on frameworks and evidence from Western societies. This raises a question whether prevalent measurement models of SEC are applicable to non-Western contexts.

This study aims to compare the measurement model of SEC between Eastern and Western countries. Using network psychometrics methods, we compared (1) dimensions underlying SEC construct, (2) sub-constructs constituting each dimension, and (3) indicators representing each sub-construct between East Asia (EA) and North America (NA).

Methods:

This study analyzed youth self-reports from the 2018 OECD Survey on Social and Emotional Skills. The study sample includes 10- and 15-year-olds (N=25,454) from two countries in EA (China and South Korea) and two countries in NA (Canada and USA).

The survey was developed based on the Big Five personality model, conceptualizing SEC as comprising five dimensions (Emotion Regulation, Task Performance, Open-Mindedness, Collaboration, and Engagement with Others) and 15 sub-constructs measured by 100 Likert-type items. Although this survey was developed considering cultural comparability, measurement non-invariance was reported across countries, warranting caution against country-level comparisons of the survey results (OECD, 2021).

Therefore, this study conducted exploratory graph analysis (EGA), a novel method estimating the number of dimensions in multivariate data using undirected network models. First, EGA and bootstrapped EGA were performed with the entire sample to ensure the integrity of dimensionality structure and robustness of item placement in each dimension. Items with item stability of <.80 or cross-loading of .70 were iteratively removed in this process. Then, this procedure was repeated with the EA and NA samples, separately, to examine cross-cultural equivalence.

Results:

Overall, the SEC measurement model of the two samples shared many similarities. Six dimensions were commonly identified with 10 sub-constructs and 51 indicators: (1) Emotion Control, (2) Optimism, (3) Task Performance (comprising Persistence, Responsibility, & Carefulness), (4) Open-Mindedness (comprising Creativity & Curiosity), (5) Prosociality (comprising Helpfulness & Empathy/Caring), and (6) Leadership.

However, some noteworthy differences were also found. First, an additional dimension of Trust was identified only in NA. Second, Sociability was a sub-construct of Prosociality only in NA, whereas Tolerance was a sub-construct of Open-Mindedness only in EA. Third, while common indicators of Emotion Control concern regulating negative emotions like anger, items related to stress were unique to NA. Also, while common indicators of Carefulness concern avoiding mistakes, items related to careful thinking/acting were unique to EA.

Conclusions:

Using EGA with large-scale international data, this study uncovered and compared the underlying structure of the multidimensional construct of SEC across Western and Eastern countries. The findings suggest that the SEC measurement model is largely comparable across two cultures, with some noted differences. This study provides a foundation for further improving international SEC assessments that are cross-culturally valid and comparable. Future studies may probe the observed East-West differences in SEC measurement to further inform culturally relevant practices promoting youth SEC in both cultures.