Scholarlink

Background <p>Chinese postgraduate students studying overseas often face academic demands and acculturative stress. The Coping Strategies Scale (CSS) was developed in China with a seven-factor structure, but coping dimensions may differ across study contexts. This study re-examined the CSS to test whether a more parsimonious structure better represents coping in Chinese postgraduates in Malaysia.</p> Methods <p>A cross-sectional online survey of Chinese postgraduate students at Universiti Teknologi Malaysia yielded 396 valid responses. The sample was randomly split for exploratory factor analysis (EFA; <i>n</i> = 200) and confirmatory factor analysis (CFA; <i>n</i> = 196). EFA (principal axis factoring with Promax rotation) was used to identify the factor structure. CFA (maximum likelihood) compared the EFA-derived five-factor model with the original seven-factor model and a single-factor model. Reliability and construct validity were examined using internal consistency, composite reliability, average variance extracted, and Fornell–Larcker. Parallel analysis supported factor retention, multi-group CFA tested configural, metric, and scalar invariance across degree level (Master vs. PhD), and HTMT provided an additional discriminant validity check.</p> Results <p>The single-factor model showed poor fit (χ²/df = 8.23, CFI = 0.542, RMSEA = 0.135). Both the five-factor and seven-factor models fit the data well. For the five-factor model, χ²(395) = 455.7, <i>p</i> = .019, χ²/df = 1.15, CFI = 0.990, TLI = 0.990, RMSEA = 0.020. The five-factor solution was slightly more parsimonious than the seven-factor solution (AIC = 655.7 vs. 672.9; ECVI = 1.66 vs. 1.70). In the full sample (<i>N</i> = 396), the five subscales showed good internal consistency (Cronbach’s α = 0.79–0.93) and adequate convergent validity (CR = 0.79–0.93; AVE = 0.56–0.61), as well as discriminant validity. Parallel analysis supported factor retention, multi-group CFA tested configural, metric, and scalar invariance across degree level (Master vs. PhD), and HTMT provided an additional discriminant validity check.</p> Conclusions <p>In this sample, the CSS was better represented by a five-factor structure than by a single-factor model, and it provided a more parsimonious summary than the original seven-factor model. This simplified structure offers a concise and valid framework for assessing coping among Chinese postgraduate students in multicultural study settings. It may support more precise coping profiles and inform targeted student support.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Psychometric validation of the Chinese Coping Strategies Scale among postgraduates in Malaysia using a five-factor model approach

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月9日

查看全文
  • Yan Hui Wang,
  • Kee Jiar Yeo,
  • Shih Hui Lee

摘要

Background

Chinese postgraduate students studying overseas often face academic demands and acculturative stress. The Coping Strategies Scale (CSS) was developed in China with a seven-factor structure, but coping dimensions may differ across study contexts. This study re-examined the CSS to test whether a more parsimonious structure better represents coping in Chinese postgraduates in Malaysia.

Methods

A cross-sectional online survey of Chinese postgraduate students at Universiti Teknologi Malaysia yielded 396 valid responses. The sample was randomly split for exploratory factor analysis (EFA; n = 200) and confirmatory factor analysis (CFA; n = 196). EFA (principal axis factoring with Promax rotation) was used to identify the factor structure. CFA (maximum likelihood) compared the EFA-derived five-factor model with the original seven-factor model and a single-factor model. Reliability and construct validity were examined using internal consistency, composite reliability, average variance extracted, and Fornell–Larcker. Parallel analysis supported factor retention, multi-group CFA tested configural, metric, and scalar invariance across degree level (Master vs. PhD), and HTMT provided an additional discriminant validity check.

Results

The single-factor model showed poor fit (χ²/df = 8.23, CFI = 0.542, RMSEA = 0.135). Both the five-factor and seven-factor models fit the data well. For the five-factor model, χ²(395) = 455.7, p = .019, χ²/df = 1.15, CFI = 0.990, TLI = 0.990, RMSEA = 0.020. The five-factor solution was slightly more parsimonious than the seven-factor solution (AIC = 655.7 vs. 672.9; ECVI = 1.66 vs. 1.70). In the full sample (N = 396), the five subscales showed good internal consistency (Cronbach’s α = 0.79–0.93) and adequate convergent validity (CR = 0.79–0.93; AVE = 0.56–0.61), as well as discriminant validity. Parallel analysis supported factor retention, multi-group CFA tested configural, metric, and scalar invariance across degree level (Master vs. PhD), and HTMT provided an additional discriminant validity check.

Conclusions

In this sample, the CSS was better represented by a five-factor structure than by a single-factor model, and it provided a more parsimonious summary than the original seven-factor model. This simplified structure offers a concise and valid framework for assessing coping among Chinese postgraduate students in multicultural study settings. It may support more precise coping profiles and inform targeted student support.