高科技產業相當仰賴策略聯盟以快速取得發展所需的資源或知識。然策略聯盟同時也為廠商帶來風險，故如何保障自身利益並建立對點夥伴的信心，向來為聯盟與廠商所關切的課題。本研究歸納出聯盟廠商的六種信心建立機制（即個人關係網絡、對夥伴之認同、目標相容性、資訊掌握、潛在經濟懲罰力與潛在法律懲罰力），以台灣資訊電子業廠商為實證對象，由高階主管進行問卷填答，總計回收146件有效之策略聯盟合作案。研究結果顯示，對夥伴之認同、目標相容性、資訊掌、潛在經濟懲罰力對點信心之建立確有幫助，但是個人關係網絡與潛在法律懲罰力為無效的機制。此外，廠商本身規模、廠商與夥伴之間的相對地位、以及來源國籍相同與否，影響到信心建立機制的組合型態。

The high technology industry depends heavily on strategic alliances to acquire resources or knowledge for growth. However, strategic alliances bring risks to firms as well. Therefore, how to get protection and build cooperation confidence when joining strategic alliances are major concerns to firms. Previous studies on cooperation protection have focused on two approaches: the transaction cost theory (emphasizing the reduction of partner’s opportunistic behaviors) and network theories (stressing the importance of trust building). However, both approaches have their limitations and tend to be struck in the myth of ethics as well as personal traits. Therefore, this study adopts another concept, confidence, to capture the essence of relationship management in strategic alliances. Confidence refers to a firm’s perceived level of certainty that its partner will pursue mutually compatible interests in an alliance, rather than act opportunistically. The conceptual differences between trust and confidence in an alliance are as follows: (1) trust implies the speculation about the partner’s intention, while confidence indicates the knowledge and control of the partner’s behaviors. (2) The more trust, the more vulnerability a firm may encounter. While the more confidence, the less vulnerability a firm may have. (3) High level of trust implies the redundancy of control. However, we propose that it is better for a firm to adopt some control mechanisms to ensure confidence. By reviewing and integrating relevant literature, we derive six major mechanisms for confidence building: personal networks, identification with the partner, goal compatibility, information completeness, potential economic punishment power, and potential legal punishment power. However, a firm might be constrained by some factors (such as the relative position, size of the firm, or the country of origin), and thus it could not adopt these confidence building mechanisms discretionarily. The information of one hundred forty-six cooperation projects were collected through structured questionnaires, filled out by high-level executives of Taiwanese firms in the information technology industry. Empirical findings suggests that: (1) the categorization of six confidence building mechanisms is showed to have good discriminant validity; (2) identification with the partner, goal compatibility, information completeness, and potential economic punishment power are helpful to build cooperation confidence, but personal networks and potential legal punishment power are not effective; (3) size of the firm, relative position between partners, and partners’ countries of origin affect the composition of confidence building mechanisms. Past researches encourage partners to develop mutual trust, but they have ignored the difficulties in developing trust, and missed the unsymmetrical problem (that is, a firm often does not worry about the trustworthiness of itself, but worry about that of its partner). Therefore, the research findings of this study could complement the literature on strategic alliances by providing clear suggestions for the reduction of cooperation risks as well as the building of cooperation confidence.