在資訊系統與軟體發展領域中，軟體流程改善(Software process improvement，簡稱為SPI)為一項重要且知識密集的工作。本研究計畫即是探討軟體組織的SPI行動所涉及的知識驅動因子並進行實證研究。具體而言，本研究匯集領域文獻中所提出的四個知識驅動因子(Knowledge-based enablers)─即Exploration、Exploitation、Absorptive capacity (AC)、與Knowledge sharing (KS)，以及在本研究團隊過去的相關研究成果基礎上，進一步提出一個針對SPI的整合型知識驅動模式、並探索與實證分析這些知識驅動因子在不同的SPI成功境界所扮演的角色與機制。對於SPI成功境界的定義，本研究擬以實務上許多軟體組織在推動SPI所使用的能力成熟度整合模式(Capability Maturity Model Integration，簡稱為CMMI)為例，以能將學術上所提出的知識因子進一步結合實務模式並交互驗證。本計畫之實證研究對象除了台灣之外、預計也包含香港與中國大陸等地區曾獲CMMI各級認證之軟體組織或企業，以進一步探討不同區域組織在研究結果上所展現之差異。本計畫預計一年完成，期間包含文獻收集、模型建立，問卷設計、調查的實施、深度探訪、研究發現的分析與探討、與研究成果的發表等重要工作。研究成果則包含模式的提出與實證、以及知識驅動因子在不同SPI境界所扮演的角色與機制等理論與實務意涵等，以提供軟體產業在推動流程改善且於邁向不同改善境界時參考。 ;Software process improvement (SPI) is a knowledge- and learning-intensive activity that is critical to the management of information system and software development. In literature, scholars have discovered several knowledge-based antecedents, i.e., exploration, exploitation, absorptive capacity (AC) and knowledge sharing (KS), that are crucial to SPI success. These antecedents have increased our understanding of how SPI knowledge is managed and operated in successful SPI implementation. However, the different roles and their comparative effects of exploration, exploitation, AC and KS in the SPI success context and how the relative effects of these antecedents are associated with different levels of progress in SPI success remain unknown. To address these gaps, this research is to develop a research model that combines the four knowledge antecedents to investigate and examine their comparative influence on successful SPI implementation. To reflect the different levels of SPI success, this research will employ capability maturity model integration (CMMI)-- an establish and widely used SPI program, and its maturities. A survey research methodology will be implemented to examine the model for CMMI-certified organizations. In order to receive a border and more generalized result, the organizational subject in the survey include not only the firms and organizations in Taiwan but also in Hong Kong and China areas. The partial least squares (PLS) technique will be used for data analysis to testify the model. This research project is proposed to conduct for a year. As the end of this research, the results will be consolidated in providing a knowledge road-map with a corresponding discussion of theoretical and practical implications for organizations to consider when pursuing superior SPI success.
