- 期刊
如何分析一致性：潛在一致模式與多項式迴歸概念與應用之簡介
How to Model Congruence: A Primer on the Latent Congruence Modeling and Polynomial Regression
摘要
一致性（congruence）是組織行為與人力資源管理的重要議題。廣義來說，學者使用一致性此概念來探討適配（fit）、相符（match）、相似（similarity）和同意（agreement）等概念。早期研究者多採用差異分數來操作一致性之概念，但因差異分數的限制，近期採用多項式迴歸（polynomial regression, PR）分析與潛在一致模型（latent congruence model, LCM）進行分析。雖然這兩類分析方式已日趨普遍，但相關中文簡介說明較少，限制我國學者使用這些技術的意願。故本文有三個目的，（1）簡介差異分數的分析方法及其限制；（2）比較PR與LCM兩種分析取向之異同；（3）以實例示範如何進行這兩類分析與解讀分析結果。最後本研究總結使用此兩類方法的實務建議，作為未來探究一致性研究議題之參考。
並列摘要
Congruence is an important research topic in organizational behavior and human resource management. Congruence comprises fit, match, similarity, and agreement between two constructs. Researchers have adopted difference scores to model congruence in the past. However, this approach is discouraged because of its potential problems. Recently, polynomial regression analysis (PR) and latent congruence model (LCM) have received considerable attention in modeling congruence. This article first reviews and discusses the limitations of the difference score approach of modeling congruence. Next, we review and compare the PR and LCM approaches. Then, we provide an illustrative example to demonstrate the uses and results of these two approaches. Finally, we summarize the uses and practical implications of these two approaches.