合併前組織地位對合併前組織認同及組織正義與合併後組織認同關聯之調節效應探討：多層次分析

Abstract

半數以上的合併案在某種程度上，最終都宣告失敗，而在合併案中多數的問題可歸咎於人為因素。然而，過去多數有關合併的研究多著重在公司策略、財務績效及合併程序，相對來說著重組織觀點的併購研究並不多。以社會認定理論觀點來看，合併對於組織的認同構成威脅，也影響員工對於組織的社會認定。而本研究主要探討合併過程中對於員工的合併後組織認同之影響因素。此外，基於多層次的研究觀點逐漸為組織行為及人力資源管理研究者重視，本研究佐以多層次分析的觀點進行研究，分別以個體層次之組織正義和合併前組織認同，與團體層次的合併前組織地位為自變數，探討兩者對於員工之合併後組織認同的影響。 研究結果顯示，在多層次分析下分配正義與互動正義對合併後組織認同存有正向影響關係，而程序正義卻未對合併後的組織認同有正向影響關係。另外，團體層次變數合併前的組織地位，僅對合併前的組織認同和互動正義與合併後組織認同之間關係產生調節效果。而合併前的組織認同與合併後的組織認同受到合併前組織地位的調節之結果，與未以多層次進行分析的相關研究結果一致。

More than half of the mergers do not fulfill the expectations set at the start to some extent. Most of the problems during the mergers could be attributed to the human factors. However, most researches concerning to mergers in the past were focused on strategies, financial performance and merger procedures; in the contrast, studies based on a organizational perspective were less. From a social identity perspective, mergers pose threat on organization identification and influence employees’ social identity toward the organizations. This study is mainly focused on the effect of organizational merger on post-merger organizational identification. Besides, multilevel analysis is paid more and more attention to the OB and HRM researchers; this study uses a multilevel analysis to investigate. The individual level factors, organizational justice and premerger organizational identification, and the group level factor, premerger organization status, are the predictors of post-merger organizational identification. The result indicates that distributive justice and interactive justice are positively related to post-merger organizational identification, but procedural justice isn’t. Furthermore, the group level factor, premerger organization status, only moderates the relationship between interactive justice and post-merger organizational identification. And the result of that the relationship between premerger organizational identification and post-merger organizational identification is moderated by premerger organization status is consistent with the result which doesn’t use in multilevel analysis.