With the growth of Web 2.0 technologies, more and more consumers share product-related information with many other consumers on various web platforms, such as product review websites (e.g., epinions.com), brand websites (e.g., dell.com), retailer websites (e.g., amazon.com), and social media. However, the question of how credible such product information is has arisen. On the Internet, the anonymous nature of consumer product reviews makes it difficult for consumers to evaluate review quality and credibility. Due to the lack of assurance of product review veracity, consumers often depend on cues to evaluate the review content accurately. One of such cues is the valence of the review: either positive or negative. Between these two valences, negative reviews have been known to be more attention getting and diagnostic than positive reviews. Thus, this study only focused on negative reviews. Another cue can be the consensus level of the review (i.e., the degree to which other consumers agree with the product opinion of the reviewer). Finally, consumers may consult the brand origin of the target brand in product reviews (either domestic or foreign brand). To date, however, little research has focused on the role of consumer ethnocentrism on brand evaluations in the context of online product reviews, despite today’s consumers often being exposed to both domestic and foreign brand product reviews in a globalized economy where major firms have multiple international branches. Thus, the purpose of this study is to examine how different cues influence consumers’ responses to negative online reviews about domestic and foreign brands. Toward that end, this study explores how a consumer’s personality trait (level of ethnocentrism) interacts with brand origin and a property of the medium (consensus information) to influence his or her brand attitude. To test the proposed hypotheses, a 2 (Brand origin: domestic versus foreign) x 2 (Consumers’ ethnocentrism: high versus low) x 2 (Consensus level: high versus low) between-subjects experiment was used. As experimental stimuli, four fictitious online retailing web pages including product reviews about a laptop computer (negative review of a domestic brand with high consensus, negative review of a domestic brand with low consensus, negative review of a foreign brand with high consensus, and negative review of a foreign brand with low consensus) were created. An Analysis of Covariance (ANCOVA) was run to test the hypothesized relationships with prior attitude toward target brand as a covariate. The results showed that there was a significant two-way interaction effect between brand origin and consumer ethnocentrism on attitude change. Consistent with predictions, when consumers’ ethnocentrism levels were high, consumers exposed to the negative review of the foreign brand exhibited significantly greater attitude change than did those exposed to the negative review of the domestic brand in a less favorable way. In contrast, when consumers’ ethnocentrism levels were low, no difference appeared between the negative review of the domestic brand and the negative review of the foreign brand in attitude change. When exposed to the negative review, attitude toward the brand was degraded for less ethnocentric consumers regardless of brand origin. Further, three-way interaction effects among brand origin, consumer ethnocentrism, and consensus level on attitude change were found. Under high-consensus condition, a negative review of the foreign brand led highly ethnocentric consumers’ brand attitudes to deteriorate, compared to a negative review of the domestic brand. Contrarily, for less ethnocentric consumers, however, there was no difference between the negative review of the domestic brand and the negative review of the foreign brand in attitude change. Under low-consensus condition, there was no difference between highly and less ethnocentric consumers in attitude change. The findings of this study suggest that the detrimental effect of negative reviews can be buffered under some circumstances. Consumers with prior attitudes toward brands tend to resist negative review of their favored brands even under high-consensus conditions. Practically, this finding implies the importance of retaining existing consumers who already have prior attitudes toward the brand. Additionally, the findings of this study suggest that consensus information, as a heuristic cue, serves as a useful indicator of review credibility. Review web sites can help consumers’ product judgments by providing consensus information. Lastly, this study used real brand names rather than fictitious brand names. Since this study takes into account consumers’ pre-existing beliefs and attitudes, its findings are applicable more easily to real-market situations than are findings regarding fictitious brand names.

伴随着WEB2.0技术的发展，越来越多的消费者在各种网络平台（例如：产品评价网站epinions.com；品牌网站 dell.com；经销商网站 amazon.com等）分享与产品有关的信息。但问题是，这些产品信息是否可信呢？互联网上，由于产品的评论者都是匿名的，这就导致消费者很难甄别评论的真假和质量。由于缺少对评论真实性的保证，消费者经常依靠一些线索来评价内容的可信度。其中的一个线索就是评论的效果：正面的还是负面的。在这两个效果间，负面效果要比正面效果更能引起关注和讨论。所以，本研究只针对负面评论。另外一个线索就是评论的共识（例如：其他人对于该评论的认可度）。最后，消费者会对品牌的来源进行追溯，看该品牌其他商品的评论。但是，到目前为止，尽管消费者都已适应全球化的时代，而且大多数企业都应开始进军海外市场，只有极少的研究针对消费者的民族中心主义对于品牌网络评价的影响。所以，本研究的目的就是要检验不同的线索对消费者针对负面评论的反应有何种影响。为了达到研究目的，笔者还将探寻消费者的个性特征（民族中心主义的强弱水平）和品牌原产国的共同影响以及媒体资源对于消费者态度的影响。为了验证假设，我们使用一个2（品牌来源国：本国vs国外）x 2（顾客民族中心主义：高vs低）的模型。作为实验样本，4个假的在线零售网站的网页（对于本国品牌带有高民族主义的负面评论，对于本国品牌带有低民族主义的负面评论，对于外国品牌带有高民族主义的负面评论；对于外国品牌带有低民族主义的负面评论）被用于刺激样本。ANCOVA被用于检验假设。结果表明产品来源国和民族中心主义对于品牌改变有着显著地双向交互作用。与预期相符，当顾客的民族中心主义高时，消费者对于外国品牌的负面评论的态度要明显强于对国内品牌的负面评论。相反，当顾客的民族中心主义低时，消费者对于国内品牌和国外品牌的负面评论所表现出的态度没有显著差异。另外，品牌来源国，消费者民族中心主义和共识水平的三项交互作用也十分明显。当共识程度高时，外国品牌的负面评论使得高民族中心主义的消费者对于品牌的态度极度恶化。相反，对于低民族中心主义的消费者而言，对于本国品牌和外国品牌的负面评论没有显著的态度区别。本文的结论建议负面评论的不良影响会受到一些情况的制约。当消费者对于之前喜爱品牌态度即使在高共识的情况下也不会受到负面评论的显著影响。针对于显示，这点表明已有消费者对于品牌态度而重要性。另外，本研究还建议达成共识的信息，作为启发式的线索，会对评论的可信度产生显著的影响。评论网站可以通过为消费者提供支持信息来帮助他们做出判断。最后，本研究使用真实的品牌而不是虚拟的品牌。所以，研究样本对于这些品牌已经有了一些态度，所以这使得本文的结论应用到现实变得更为容易。