許多組織間學習文獻強調，廠商藉由組織間關係 (inter-organization relationships, 以下簡稱 IORs ) 的互動，是進行組織間學習的重要途徑；不過，這些文獻並未建構一個能使知識有效率移轉的理論模型，特別是在廠商無法監測對手的私人訊號時，該如何擬定知識移轉策略始終付之闕如。因此，藉由賽局理論，找出私人監測架構下的均衡知識移轉策略，將是知識移轉研究建立理論依據的重要課題。本文利用重複賽局架構，建構IORs 的知識移轉模型。模型中，我們根據 Matsushima(2004) 的「回顧階段」建構廠商在私人監測下的知識移轉之策略。本文發現：(一)當對手未通過檢視門檻時，廠商可以利用混合行動而非處罰對手，避免彼此陷入皆不合作的不利結果。(二)回顧策略為長期互動的序貫均衡，廠商可獲得知識移轉的個人理性及效率效用。

The studies of inter-organization learning had emphasized that firms could learn knowledge and develop technology through by interorganization relationships (hereafter IORs). However, these literatures had not
constructed a theoretic model under which made knowledge transfer efficiently. In specialty, these literatures failed to find the equilibrium strategy for knowledge transfer in IORs with private signals. Therefore, by the game theory, how to design the knowledge transfer
strategies with private monitoring is the most important issue for the studies of inter-organization learning.In this paper, the repeated game theory is applied to establish a kind of IORs knowledge transfer with
private monitoring. According to Matsushima (2004), by the review phase, the firms will regularly examine the opponents whether pass the threshold or not and as the reference of adjusting action to the next period. The
results show that (1) when the opponents do not pass the threshold, the firms can use the mixing action but not punishment to avoid the unfavorable results from mutually incorporation. (2) Review strategy is the sequential equilibrium in our model and the firms in IORs can get the individually rational and efficient utility