A Meta-analytic Investigation of the Role of Valence in Online Reviews
通过元分析，研究了在线评论的情感倾向（正面或负面）对感知有用性和产品态度的不同影响，发现负面为主但含少量正面的评论对感知有用性影响最大，而正面为主且无负面评论对态度影响最强，且效果因产品类型和品牌熟悉度而异。
Based on a meta-analysis, this study aims to clarify the influence of online review valence (i.e., whether reviews in a review set are predominantly positive or negative) on various dependent variables, in particular on perceived usefulness of the reviews and on attitudes toward the product. The findings suggest that review valence affects perceived usefulness in a different way than it affects attitudes. We find an optimum impact level for perceived usefulness and a ceiling effect for attitudes. Primarily negative review sets that include few positive reviews show the strongest effect on perceived usefulness, while positive review sets that include few or no negative reviews are most successful in influencing attitudes. Additionally, review valence has a stronger influence on perceived usefulness when the reviews refer to experience (rather than search) products, and it has a stronger influence on attitudes for unfamiliar (rather than familiar) brands. Finally, the strongest influence of review valence refers to recommendation intentions. These findings provide several implications for researchers and practitioners.