Investigating reciprocal causation in organizational behavior research
探讨了检验非递归因果模型时研究者应考虑的问题，包括如何指定模型以减少估计偏差并便于整体解释，并通过一系列分析证明了这些问题的重要性。
Abstract This paper examines issues that should be considered by researchers interested in testing nonrecursive causal models. It identifies choices that can be made in specifying those models which will minimize bias in estimation as well as facilitate the overall interpretation. A series of analyses demonstrate the importance of these issues. Implications for organizational behavior research and recommendations are discussed.