On the Use of Spline Regression in the Study of Congruence in Organizational Research
针对组织研究中一致性测量使用绝对差分的缺陷，本文提出基于样条回归的替代方法，通过实证示例证明其优于分段回归，为研究者提供更可靠的假设检验工具。
The study of congruence is central to organizational research. Congruence refers to the fit, match, similarity, or agreement between two constructs and is typically framed as a predictor of outcomes relevant to individuals and organizations. Previous studies often operationalized congruence as the algebraic, absolute, or squared difference between two component variables. Difference scores suffer from numerous methodological problems, which stimulated the development of alternative procedures. For algebraic and squared difference scores, the primary alternatives involve linear and quadratic regression equations. For absolute difference scores, the extant alternative is piecewise regression, which avoids certain problems with absolute difference scores but relies on untested assumptions that are central to congruence research. In this article, we develop an alternative to absolute difference scores based on spline regression, yielding a comprehensive approach for testing hypotheses that underlie absolute difference scores while avoiding the shortcomings of piecewise regression analysis. We demonstrate the advantages of spline regression over absolute difference scores and piecewise regression using an empirical example.