本論文主要探討愉悅的嗅覺刺激對產品評價與資訊處理的影響。透過兩個子研究分別瞭解產品嗅覺刺激與情境嗅覺刺激各自的影響，以推敲可能性模型做為貫穿全文的主要理論模型，並從多元角色觀點與享樂權宜觀點建構整合模型。產品嗅覺刺激指與產品本身直接有關的氣味，過去在此議題上的研究缺口有二，分別為未考量區辨氣味與產品搭配適合性所需具備之動機，以及不同產品類型之調節效果，例如，氣味對於家用清潔品及保養品的診斷性及攸關性皆較高，但對於筆記型電腦、手機、手錶或隨身碟等產品之診斷價值較低。情境嗅覺刺激指存在於情境中但與被評估之標的產品無關的氣味，過去缺乏探討情境嗅覺刺激影響資訊處理模式之相關研究，但此議題無論對學術或實務都相當重要，因為許多研究人員皆不斷地想探索那些因素會影響人們的資訊處理模式。而且，行銷人員總是希望能藉由產品資訊說服消費者，尤其當不知名品牌的新產品問世時，更仰賴消費者對產品資訊的仔細處理而被說服，進而產生較佳評價。因此，若能進一步瞭解情境嗅覺刺激對資訊處理的影響，對行銷人員在氣味氛圍的營造與操作上，將可提供可參考的方向。 本論文透過兩個子研究並採用實驗法，分別瞭解產品嗅覺刺激與情境嗅覺刺激的影響。研究一又可分為兩個主題，主題一採3 (產品嗅覺刺激：無vs.適合性vs.不適合) * 2 (嗅覺刺激診斷性：高vs.低) *2 (涉入程度：高vs.低)之12組受測者間實驗設計。主題二為情境嗅覺刺激之前導研究，採3 (情境嗅覺刺激：無vs.適合性vs.不適合)之3組受測者間實驗設計。研究二採2 (情境嗅覺刺激的提供：有vs.無) * 2 (廣告訴求：享樂vs.功利) *2 (產品論點品質：強vs.弱)之8組受測者間實驗設計。 關於產品嗅覺刺激有三點重要發現，分別為：(1) 對低資訊處理涉入者來說，無論嗅覺刺激是否具診斷性，氣味愉悅性都是主要的影響因素，此時，產品嗅覺刺激乃為情感性角色，因任務結果對低資訊處理涉入者並不攸關且不感興趣，在這種情況下，其易採取週邊說服路徑，視產品嗅覺刺激為週邊線索，只要氣味具愉悅性，雖適合性不高，仍可提升產品評價。(2) 對高資訊處理涉入者來說，嗅覺刺激具診斷性時，產品嗅覺刺激主要扮演認知性角色，因任務結果對高涉入者為攸關且感興趣，在此情況下，其易採取中央說服路徑，視該氣味為攸關判斷的產品論點，所以會仔細推敲該嗅覺刺激與產品的關係，以決定搭配適合性，對於適合的嗅覺刺激評價會比不適合的佳，此時，嗅覺刺激適合性就類似於強弱產品論點的角色。(3) 當嗅覺刺激不具診斷性時，其認知性角色就會消失，此時，高資訊處理涉入者仍然採取中央說服路徑，但氣味已被視為與判斷無關的週邊線索，所以高涉入者並不會再仔細推敲該嗅覺刺激與產品的關係，適合與否也就不會產生差異，甚至高涉入者還會啟動修正機制，排除氣味愉悅性的影響。 在情境嗅覺刺激部份亦有四點重要發現，分別為：(1) 當愉悅的情境嗅覺刺激存在時，會有助於增加知覺者的正向心情，進而增加消費者對情境標的產品之資訊處理程度並促進其採取中央資訊處理路徑，對於強勢論點品質之產品評價會優於弱勢論點。(2) 愉悅的情境嗅覺刺激也會改變消費者對不同產品資訊類型的偏重，即在資訊處理或編碼階段會對享樂資訊特別關注，因此，在事後對於享樂資訊的提取比率較高，上述效果在享樂廣告訴求下更為明顯。(3) 有愉悅的情境嗅覺刺激存在時，會讓消費者對享樂訴求廣告之產品評價優於功利訴求。(4) 本研究進一步建構情境嗅覺刺激影響產品評價之多元關係，結果發現當產品類型較偏功利性時，則情境嗅覺刺激會先經由引發好心情再增加資訊處理程度，最後才影響產品評價，其中正向心情與資訊處理程度在模式中都是部份中介的角色，當產品類型較偏享樂性時，則情境嗅覺刺激影響產品評價就完全經由心情機制。 / This study adopts a multiple-role view from the elaboration likelihood model (ELM) and hedonic contingency model to explore the influences of product scents and ambient scents. The product scents are scents that consumers smell directly from products. In the literature on product scents, there are no research studies considering the moderate role of product types and the motivation to recognize scents appropriately. Ambient scents are scents smelled that are unconnected to products. In the literature on ambient scents, there are no research studies considering how ambient scents can influence information processes. However, many scholars want to explore the factors that can influence information processes. Furthermore, a new product launch by an unknown brand can convince consumers depending on consumers’ careful processing of product information. Thus, this issue is very important to academic study and practice. This study adopts two experiments that realize the effects of product scents and ambient scents separately. Study One can be divided into two parts. Part A adopts a 3 (product scents: none, fit, unfit) * 2 (involvement: high vs. low) * 2 (diagnostic of scents: high vs. low) between-subjects design. Part B is a pilot study of ambient scents that adopts a between-subjects design with 3 factors (ambient scents: none, fit, unfit). The Study Two adopts a 2 (ambient scents: yes vs. no) * 2 (advertisement appeal: hedonic vs. utilitarian) * 2 (argument quality: strong vs. weak) between-subjects design. The results of Study One have three points: (1) for consumers with low involvement in tasks, whether scents are diagnostic for products or not, pleasant scents have significant effects. Because people who are involved in tasks at a low level and are not concerned or interested in outcomes, they are almost persuaded by peripheral routes and regard product scents as peripheral cues. Even though the scents do not fit with the products, they still can cause positive product evaluation depending on the nature of the pleasure. For people with low involvement in tasks, product scents mainly play an emotional role. (2) For people who are highly involved in tasks and really concerned or interested in outcomes, they are almost persuaded by central routes and regard product scents as arguments, especially when scents are diagnostic for products. For them, product scents not only play an emotional role but also a cognitive role. They will evaluate the similarities or differences between products and scents. A scent that fits with a product will cause a more positive product evaluation than a scent that does not fit a product. (3) For consumers who are highly involved in tasks, as scents are not diagnostic for products, the cognitive role will disappear. The scents are regarded irrelevant to product evaluation, so consumers do not evaluate the similarities or differences between products and scents. Whether a scent fits a product or does not is unimportant for them; furthermore, they will correct for and eliminate the effects of pleasant scents in product evaluation. The results of Study Two have four points: (1) when pleasant ambient scents can be provided by marketers, they can cause a positive mood among perceivers and enhance the level of information processing toward the focal product within the context. Furthermore, pleasant ambient scents can encourage consumers to adopt central routes to process information. People will evaluate products with strong argument quality more positively than products with weak argument quality. (2) Pleasant ambient scents also draw people’s attention toward different information types. People will attend more to hedonic product information, especially with hedonic advertisement appeal. (3) A pleasant ambient scent that matches a hedonic advertisement appeal can cause a more positive product evaluation than a pleasant ambient scent that matches a utilitarian advertisement appeal. (4) For a utilitarian product type, the effects of pleasant ambient scents on product evaluation will cause a positive mood and enhance the level of information processing. Mood and cognition play a partial mediation role. For a hedonic product type, the effect of pleasant ambient scents on product evaluation is fully mediated by a positive mood.