Intentional self-regulation in middle adolescence: An ESEM examination on a diverse Taiwanese sample

• In Event: Factors Affecting Goal-Directed Behavior

Abstract

Intentional self-regulation (ISR), pertinent to goal management and attainment (Gestsdóttir & Lerner, 2007), is an important indicator of healthy and adaptive functioning in adolescence and adulthood (e.g., Mueller et al., 2011; Ziegelman & Lippke, 2007). Most research conceptualizes ISR using the selection, optimization, and compensation (SOC) model. The tripartite structure of SOC model has been empirically validated in adulthood (Freund & Baltes, 2002) but is not yet apparent during adolescence (Geldhof et al., 2015). Because ISR is still maturing in adolescence, it is possible that adolescents do not differentiate selection, optimization, and compensation in their goal-directed thoughts and actions. For example, when adolescents ‘think about their plan,’ their thinking might indicate that they are committed (selection) and they try to optimize their resources to realize their goals (optimization). Any single item measuring SOC might therefore indicate multiple SOC factors. Because a confirmatory factor analysis (CFA) restricts cross loadings between items and factors, CFA might not be ideal for the analysis of adolescents’ ISR. An exploratory structural equation model (ESEM) approach, on the other hand, allows all items to load onto all factors. Therefore, ESEM might afford a more developmental fitting examination on adolescents’ undifferentiated ISR.
This study followed a diverse group of adolescents in Taiwan for three years (63% Han Taiwanese, 37% Indigenous Taiwanese, 51% male). The participants’ average age was 12.48 years during Wave 1, and it was 13.43 and 14.36 years, respectively, at Waves 2 and 3. ISR was measured using a short Mandarin version of the SOC scale (18 items, Gestsdóttir & Lerner, 2007). All items were positively stated with Likert-style response options. The translation of the SOC measure followed the procedure of translation and back translation. The ESEM approach was used to analyze the SOC data collected at Waves 1, 2, and 3. Longitudinal (Waves 1, 2, and 3) and multi-group (male vs. female, Han Taiwanese vs. Indigenous Taiwanese) tests of invariance of the ESEM models were also conducted.
The results show that the ESEM approach afforded better model fit and meaningfully smaller factor correlations as opposed to the CFA approach (see Table 1; Wave 2 results are presented as an example). Items that were consistently loaded on the SOC factors across waves are different from those identified in the Northern American and European samples (Gestsdóttir & Lerner, 2007; Gestsdóttir et al., 2015, see Table 2). Longitudinal invariance was established between the Wave 2 and Wave 3 ESEM models, but not with the Wave 1 model. Gender and ethnic-group invariance was achieved for models on all three waves, though some item intercepts were required to be freely estimated.
Our findings show that the process of ISR differentiation might begin at the ages of 14 or 15. However, clear differentiation is not evident during middle adolescence. In addition, ISR development seems to be universal across gender and ethnic groups in our sample. Importantly, contextual factors, such as cultural norms or the educational system, might determine which goal-directed behavior are more present in the individual.

Authors

• PEI-JUNG YANG, National Chengchi University
  Presenting Author

• G. John Geldhof, Oregon State University - Corvallis

• Dian Yu, Oklahoma State University - Stillwater

• Yue Ni, Oregon State University - Corvallis