Poster #64 - Psychosocial Development in Late Adolescence: A Cross-Cultural Perspective

• In Event: Poster Session 19

Abstract

Background and Purpose:
Common aspects of psychosocial development have long been assumed universally applicable yet recognized and studied as nuanced. Scores of studies surrounding age variations and gender differences have informed our understanding of psychosocial developmental patterns. Less has been done to understand the complexity of cultural differences in accepting the assumption of ubiquitous stage-like progression and application to mental health outcomes across cultures. The purpose of this study was to examine the cultural relevance of common psychosocial measures and explore transitional patterns in, and predictive outcomes of, psychosocial development among late adolescents across cultures.
Method: A sample of 1,335 late adolescents (372 Taiwanese, 365 Italian, and 598 from the US) provided responses for their own psychosocial development, depression, and self-worth. Several single factor confirmatory factor analyses were conducted to evaluate whether a given latent construct properly loaded the same indicators across cultures for setting up a configural invariance model for further measurement invariance tests. Once configural measurement invariance was established across cultural groups, multiple-group confirmatory factor analysis with alignment optimization was applied to identify indicator intercepts and factor loadings that were non-invariant across cultural groups. Latent profile analysis (LPA) was employed to identify differential psychosocial developmental patterns based on z-scores of four cognitive autonomy scales, four identify scales, and the shyness independently for country. The final number of obtained classes was based on commonly accepted model-fit statistics. The optimal number of LPA classes was then regressed on self-worth and depression while controlling for age, gender, parental education, and family income.
Results: Configure invariance was obtained for modified versions of Cognitive Autonomy subscales including Evaluative Thinking, Voicing Opinions, Making Decisions, and Self-Assessing; Identity Subscales, Achievement, Moratorium, Foreclosure, and Diffusion; for Shyness; and for Self-Worth. The models (except for the saturated model), CFI values ranged from .973 to .996, RMSEA values ranged from .029 to .050, and SRMR values ranged from .019 to .039. Results from the alignment optimization analysis further showed that there were some differences in item intercepts and threshold. Parameters invariance percentage ranged from 2.8% to 13.3%. Latent profile analyses from 2-class to 6-class models were conducted for each country. Both Italy and Taiwan sample derived a 4-class solution whereas US sample had a better fit for 5-class solution.
Though transitional patterns in psychosocial development varied greatly among late adolescence across counties, in general, Class 1 represented the least desirable status whereas Class 4 (or Class 5) represented the most desirable status for each country. Class membership predicted mean scores in self-worth and depression. Class 5 in US, Class 4 in Italy, and Class 3 in Taiwan had the best mental health status, whereas Class 1 in US, Class 3 in Italy, and Class 1 in Taiwan did not fare as well.
Conclusions and Implications: Results implicated there were measurement non-invariance across countries and therefore when evaluate psychosocial development across culture, there is a need to use a more culturally sensitive approach to understand the extent to which development is related to healthy or risk outcomes.

Authors

• Troy Beckert, Utah State University

• ChienTi Plummer Lee, University of Wisconsin - Milwaukee
  Presenting Author

• Paolo Albiero, University of Padua