The impact of online consumer reviews on online purchase decisions has increased with the growth of e-commerce. This paper tries to explain how rating dispersion impacts the process of review consumption based on the heuristic systematic model. For this research, 10,198 online consumer reviews for 516 DVD products were collected from Amazon.com using a web data crawler. Our results show that when trusting average ratings (i.e., when rating dispersion is low), the incentive to read individual reviews decreases because of the principle of least effort. In this case, consumers consider average ratings to be representative of collective intelligence, so rating inconsistency negatively impacts review helpfulness. On the other hand, when average ratings are not trusted (i.e., when rating dispersion is high), the incentive to read individual reviews increases because of the principle of sufficiency. When it happens, extreme ratings affects review helpfulness more because extreme opinions are not ambiguous. Our findings provide new perspectives to address the inconsistent findings of the previous studies on rating and review helpfulness, and the practical implications for e-commerce platforms.

随着电子商务的发展，在线消费者评论对在线购买决策的影响已经增加。本文试图基于启发式系统模型来解释评分差异如何影响评论消费过程。在这项研究中，使用Web数据搜寻器从Amazon.com收集了516个DVD产品的10,198个在线消费者评论。我们的结果表明，当信任平均评分时（即评分分散性较低时），由于省力原则，阅读个人评论的动机降低了。在这种情况下，消费者将平均评分视为集体智慧的代表，因此评分不一致会对审核帮助产生负面影响。另一方面，如果不信任平均评分（即评分分散性很高），由于充分性原则，增加阅读个人评论的动机。发生这种情况时，由于极端意见并不明确，因此极端评分会更影响评论的有用性。我们的研究结果提供了新的观点，以解决先前关于评级和评论有用性的研究不一致的发现，以及对电子商务平台的实际影响。