企業資訊系統採用：基於TOE模型之整合性分析

Organizational Adoption of Information Technology: A Meta-Analysis Based on the TOE Model

企業對於資訊系統的採用是一個重要的研究議題，然而在過去相關的研究，對於可能影響因素及所產生的效果，彼此間有一定程度的分歧，因此本研究預期透過整合分析(meta analysis)彙整企業資訊採用文獻，而彙整的架構則使用整合各方觀點的TOE框架，完整將過往文獻所發現的影響因素依照科技、組織以及環境三構面加以統整，並透過整合分析的統計分析得出各因素在考量不同研究後的匯總效果，尤其所採用的隨機整合模型(random effect model)，對於估計出的匯總效果，具有外推到未知樣本(如：未收集或未來可能相關研究)的效力。本研究預期提出綜整性的企業資訊採用模型，嚴謹彙整出過去文獻顯著且一致的影響因子，更重要的是探討各因素在不同情境下的差異，顯示企業資訊系統採用的權變性質。

The organizational adoption of information technology (IT) has been an important research area for decades. A number of factors have been investigated, but some previous findings about their effects are inconsistent and need further consolidation. To address this need, this study aggregates previous findings using the meta-analysis. The TOE model is a major framework that includes factors associated with technology, organization, and environment, and it is comprehensive enough to absorb various variables examined in literature. Meta-analysis allows us to aggregate previous findings statistically to draw more robust conclusions regarding their effects. The random effect model can extend in-sample estimates to out-of-sample generalizability. While knowledge of some factors suffers from insufficient study and/or inconsistent findings, we identify factors that are significant and stable across studies. Furthermore, direction and strength of a factor is distinct across studies. We use the moderation analysis to show that effects of factors are contingent on the three contextual factors. Overall, we propose an integrated model for organizational IT adoption. The TOE framework provides the comprehensive dimensions, and 14 factors are identified with rigorous support of the meta-analysis. More importantly, the moderation analysis explores the impacts of contextual factors, relationships that may not necessarily be apparent from individual studies, leading to an extended TOE framework.