Scholarlink

<p>Common method variance (CMV) has been receiving great concern in the measurement of constructs because CMV might potentially bias research findings. Common method biases (CMB) in parameter estimates are the biases resulting from CMV. However, the presence of CMV does not necessarily cause significant CMB. CMV correction is needed only when common method biases are detected to be significant. The traditional chi-square difference test is commonly used for CMB detection in the literature, but it not sensitive to the amount of CMV. In this study, the Bollen-Stine bootstrap method is applied to the chi-square difference test for CMB in estimated trait correlations based on the unmeasured latent method construct (ULMC) model. Simulation results have shown that the bootstrap-adjusted chi-square difference test is more sensitive and demonstrates greater ability to test for CMB than the traditional chi-square difference test. The bootstrap-adjusted chi-square difference test has been illustrated with a simulated dataset. We suggest that the test be used to detect CMB for empirical studies facing the threat of CMV.</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

An improvement in the detection of common method biases

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年7月14日

查看全文
  • Chien-Fan Chen,
  • Cherng G. Ding

摘要

Common method variance (CMV) has been receiving great concern in the measurement of constructs because CMV might potentially bias research findings. Common method biases (CMB) in parameter estimates are the biases resulting from CMV. However, the presence of CMV does not necessarily cause significant CMB. CMV correction is needed only when common method biases are detected to be significant. The traditional chi-square difference test is commonly used for CMB detection in the literature, but it not sensitive to the amount of CMV. In this study, the Bollen-Stine bootstrap method is applied to the chi-square difference test for CMB in estimated trait correlations based on the unmeasured latent method construct (ULMC) model. Simulation results have shown that the bootstrap-adjusted chi-square difference test is more sensitive and demonstrates greater ability to test for CMB than the traditional chi-square difference test. The bootstrap-adjusted chi-square difference test has been illustrated with a simulated dataset. We suggest that the test be used to detect CMB for empirical studies facing the threat of CMV.