口碑是一種社會行為。因e化時代的來臨，口碑不再侷限於面對面之溝通；相反地，應該廣泛地擴及於電子載具之上。本研究檢視口碑在不同載具群組時對消費者行為之影響，及與其他個人因素是否有交互作用之存在。本研究亦探討正負口碑下的順序效應，結果發現: 一、受測者在負口碑下較易受聲音傳輸之影響，如面對面及Skype口碑， 二、在訊息整合時，無論載具為何，後口碑皆會減弱前口碑，近因效應因而產生， 三、訊息兩極且不分前後順序時，電子口碑較傳統口碑易受負面訊息影響， 四、載具與順序、載具與個人因素及順序與個人因素之交互作用皆未發生，可推因為載具與順序效應各自獨立，本身已有很強之主效果。 此為第一個針對口碑互斥屬性單獨探討之研究，不與廣告或試用經驗混談；同時也是第一份綜合討論起首效應、近因效應，及與心理帳戶之損失趨避關聯性之報告。無證據顯示在順序效應下，起首效應會發生。另外發現後口碑應給予較高之權重，並支持負向效果；即後負口碑引發負面消費者行為較前正口碑引發正面消費者行為為強。期望理論價函數之損失趨避係數亦延伸引用於此。 未來研究可針對近因效應與損失趨避之加乘進行探討，如後負口碑之最高係數能否膨脹到4.84，解釋該口碑在消費者行為上扮演關鍵性貶抑角色之原因。 ;Word-of-Mouth (WOM) is a social behavior. With the advent of internet-era, WOM can impossibly be limited to face to face (FTF) communication; rather, shall be largely engaged in electronic vehicles. This Study examines how (e)WOMs in different vehicles/groups would influence customer behaviors and how they interact with other individual variables. Also investigated is the order effect for integrating information. Suggestions from the findings are (1) Subjects exposed to negative information are easily influenced by FTF and Skype that are in form of sound, (2) In integrating information, the later (e)WOMs can lessen the former one in any vehicle; that is, the recency effect stages, (3) Under uncertainty and regardless of the order, the eWOMs exhibit more susceptibility to the negative messages in comparison with the traditional FTF WOM, and (4) An apparent departure from interactions effects generated in the vehicle type and the order type, the vehicle type and the moderators, and the order type and the moderators. The sought explanation is that the vehicle type and order type are independent of each other and with strong main effects themselves . This is the first study to examine (e)WOMs integration process which deserves of research of its own, rather than combined with advertising or trial. The paper is also the first to investigate the combined effects of primacy and recency along with its connection to the loss aversion in mental accounting. No evidence of primacy effect is inferred. The findings support the later inconsistent information to be given increased weight and the negativity effect as well. In sum, later negative messages are to generate the strongest negative behavioral intention than former positive messages. The coefficient of the loss aversion in Value Function Prospect Theory is extended to apply to this. The future research could direct to the recency effect multiplied by the loss aversion which makes the later negative (e)WOMs swell into 4.84 at the highest and accounts for the later negative (e)WOMs to play a significantly detrimental role in determining the consumer behaviors; therefore.