기존 온라인 구전 관련 연구에서는 온라인에서 발생하는 구전의 내용을 긍정적인 구전과 부정적인 구전으로 구분 하여 각각의 영향력에 대해 연구가 진행되어왔다. 하지만 현실적으로 온라인에서는 동일한 대상에 대해 긍정 혹 은 부정의 한 쪽으로 치우친 일방적인 내용의 구전보다는 긍정적인 구전과 부정적인 구전이 함께 포함된 경우가 많다. 이러한 현실과는 달리 동일 대상에 대해 긍정과 부정의 구전이 같이 제시되는 상황에 대한 실증연구는 아 직 이뤄지지 않고 있다. 본 연구는 기존 연구와 달리 온라인에서 긍정적인 내용과 부정적인 내용이 함께 제공되 는 양면구전(two-sided word-of-mouth)을 제시하고, 일면구전(one-sided word-of-mouth) 대비 양면구전 이 소비자들에게 어떤 영향을 줄 수 있는지를 확인하는데 목적이 있다. 비대면적 상호작용을 하는 온라인 구전의 특성으로 인해 온라인 구전은 구전 효과에 중요한 영향을 미치는 사 회적 유대감(social tie)과 구전에 대한 신뢰성(credibility)이 오프라인 대비 상대적으로 낮게 형성되는 단점 역시 존재한다. 본 연구에서는 이러한 온라인 구전의 낮은 신뢰성을 극복하기 위한 한 방안으로 양면구전의 효과를 살펴보고, 추가적인 조절변수들의 영향을 같이 고려함으로써 양면구전의 효과가 더욱 커지는 상황에 대해 연 구하고자 한다. 구체적으로 본 연구는 온라인에서 양면구전이 구전내용에 대한 신뢰성의 증가를 매개로 하여 소 비자의 브랜드 태도에 영향을 주는지를 확인하고, 양면구전의 효과를 조절하는 변수로서 소비자의 사전지식과 브랜드 명성의 영향을 살펴보았다. 본 연구의 결과는 다음과 같다. 첫째, 양면구전이 일면구전보다 소비자의 브랜드 태도가 더 높게 형성된다는 것을 확인할 수 있었다. 또한 양면구전은 일면구전과 비교할 때 소비자가 느끼는 구전에 대한 신뢰성을 높이는 역할을 하는 것을 확인하였다. 추가 분석을 통해 양면구전과 일면구전의 효과 차이는 신뢰성에 의해 매개되고 있 음을 확인할 수 있었다. 둘째, 양면구전에 포함된 부정적 정보의 강도에 따라 이러한 효과가 다르게 나타날 수 있는데 양면구전에 약한 부정적인 메시지가 포함된 경우에는 일면구전보다 더 높은 구전효과가 나타났지만, 양면 구전에 심각한 부정적인 메시지가 포함된 경우에는 일면구전보다 더 낮은 구전효과가 나타났다. 셋째, 양면구전 의 영향력은 소비자의 사전지식에 의해 조절되는 것으로 나타났다. 소비자의 사전지식이 낮은 경우에는 양면구전 의 긍정적인 효과가 나타났지만, 소비자의 사전지식이 높은 경우에는 일면구전 대비 양면구전의 차별적 효과가 나타나지 않았다. 마지막으로 양면구전의 영향력에 대한 브랜드 명성의 조절적 역할을 예상했지만 실증 결과 이 는 나타나지 않았다. 끝으로, 본 연구를 통해 얻을 수 있는 이론적 및 실무적 시사점을 살펴보았으며, 한계점 및 향후 연구방안에 대해 논의하였다.
The existing research on online-WOM (Word-Of-Mouth) has only looked into either negative or positive WOM, not the combined type of WOM messages including both positive and negative messages. In the real world, however, we can easily find mixed WOM messages that include both positive and negative informations about a product or service. We conceptualize such mixed WOM messages as ``two-sided WOM.`` In this research, the authors investigate the influence of two-sided WOM on brand attitudes. There are a number of differences between online- and offline-WOM. One of the major differences is the importance of credibility from a receiver`s point of view. In the offline contexts, WOM usually involves face-to-face communications between sender and receiver. In the online contexts, however, people use computer-mediated communications for WOM, so they don`t directly know who sends the messages. Such anonymity in the online WOM leads to increased lack of message credibility compared to the off-line WOM. The credibility of WOM is an important factor for enhancing the effectiveness of WOM. In this research, the authors suggest two-sided WOM as a means to increase the credibility of WOM, and it also results in increased the brand attitudes. Two-sided WOM has a similar psychological mechanism to two-sided advertising. The negative message in two-sided ad enhances source credibility, and it subsequently increases consumers` advertising acceptance. In a similar vein, the negative message in two-sided WOM increases the credibility of online WOM and thus the influences of WOM messages on consumers` evaluation. A prior research also suggests that negative WOM is more credible than positive WOM in online context (Chatterjee 2001). This research examined how two-sided WOM influences the brand attitudes and found out that the credibility of WOM plays a mediating role between two-sided WOM and consumer evaluations. In addition, the authors also examined how brand reputation and prior knowledge moderate this effect. Results of the study indicate that two-sided WOM increases the credibility of WOM relative to one- side WOM (Mtwo-sided = 4.51 vs. Mone-sided = 3.70 ; F(1,207) = 22.31, p < .001), and that the consumers` brand attitudes are more favorable with two-sided WOM (Mtwo-sided = 4.54) than with one-sided WOM (Mone-sided = 4.05), except for a situation when two-sided WOM includes severe negative messages (Msevere = 3.45) in it (F (2,330) = 29.761, p < .001). And we also examine the credibility of WOM has the mediation role between WOM type and brand attitudes. In addition, consumers` prior knowledge about the target product moderates this effect (F (1,203) = 6.988, p < .05). When consumers` prior knowledge is high, there is no significant difference between one- (Mone-sided = 4.21) and two-sided (Mtwo-sided = 4.19) WOM in terms of brand attitudes. However when consumers` prior knowledge is low, two-sided WOM (Mtwo-sided = 4.76) is more effective than one-sided WOM (Mone-sided = 3.95). Against our expectation, the authors couldn`t find the moderating effect of brand reputation. We anticipate that two-sided WOM influences on brand attitude when only brand reputation is low. The result, however, suggests that there is no difference between the levels of brand reputation (p > .05). We analysed the reasons why brand reputation couldn`t moderate the effect of two-sided WOM. First, the credibility of two-sided WOM may be more important factor on brand attitude compared to brand reputation. Second, the problem can be made by our selection of brands. Finally, the result of ad hoc test says there is the gender difference on influence of two-sided WOM. Consistent with our expectation, we found that brand reputation moderates the influence of two-sided WOM in male group (partially supported; p = .10), whereas reversed effect of brand reputation on two-sided WOM was examined in female group. The study has the theoretical and practical implications. First, this research describes the influence of mixed-valence WOM (two-sided WOM). Prior researches are interested in the effect of valences (i.e., positive or negative WOM), this study introduces the mix-valence message. And we find two-sided WOM makes that consumers perceive the message as credible, consequently increases brand attitude. Second, marketers should be interested in not only positive message but also negative. This study suggests negative message can increase the credibility of WOM message, thus marketers also manage negative WOM as the means for enhancing the credibility of WOM. Third, marketers should consider the credibility of WOM. Generally, marketers focus on making good message only, but the credibility of WOM is very important for changing consumer`s attitude. Thus enhancing the credibility of WOM should be strategically considered. Finally, the target of WOM strategy should be the naive consumer group. This study suggests the expert group was less influenced than the naive group. There are some limitations and guidelines for future studies. First, empirical study was conducted in experimental condition. Thus this result has limited external validity. Second, we only selected digital cameras for experiment. For increasing internal validity, next study should conduct with other products. Third, this study focuses on replied message on products, but there are many types of online WOM (e.g., comments on the blog, consumer`s experience story etc.). Future research on two-sided WOM should consider different kind of online WOM. Fourth, this study manipulated the message strength of negative WOM through varying the importance of attribute. Next study can manipulate the message strength of WOM by changing the number of negative messages or using different expressions. Finally, we can`t suggest how company control the two-sided WOM for their brand attitude. Thus future research can consider it.