默认好评对消费者购买行为的影响  

Impact of Default Good Reviews on Consumer Purchase Behavior

作　　者：赵英男 王欣 王全胜[1] 闵超[2] 陈晓威[2] ZHAO Yingnan;WANG Xin;WANG Quansheng;MIN Chao;CHEN Xiaowei(School of Business,Nanjing University,Nanjing 210093,China;School of Information Management,Nanjing University,Nanjing 210093,China)

机构地区：[1]南京大学商学院,南京210093 [2]南京大学信息管理学院,南京210093

出　　处：《管理科学》2020年第4期137-148,共12页Journal of Management Science

基　　金：国家自然科学基金(71872082,71372035)。

摘　　要：在网络购物情景下,评论操控成为一种普遍的现象。鉴于消费者评论的缺乏,默认好评成为一些在线购物平台操控评论的手段。不同于一般的在线评论操控,默认好评具有商家不可控和消费者易识别的特征,已有的评论操控研究尚未关注这种特殊评论操控的影响和作用机制。通过爬取中国电商平台的真实产品评价和销售数据,形成面板数据并基于此构建双向固定效应模型,探讨并检验默认好评数量对消费者购买行为的影响,同时从产品层面(产品类型和流行度)和商家层面(经营模式)讨论影响的边界条件;此外,结合实验法检验产品诊断性的中介作用,实验采用单因素的组间设计,通过操控产品评论页呈现的默认好评数量形成4个样本组,共有234名学生参与研究。上述研究设计有助于发挥二手数据较高的外部效度和实验方法因果推断的双重优势。研究结果表明,①默认好评数量与产品销量呈负相关,当默认好评数量达到一定限度时,产品购买显著减少;②产品类型、流行度和经营模式的调节作用显著,默认好评对搜索型产品(与体验型产品相比)、流行度低的产品(与流行度高的产品相比)或经销商销售的产品(与供应商直销的产品相比)负向影响更大;③消费者感知产品诊断性完全中介默认好评对消费者购买意愿的影响,默认好评数量的增加降低产品诊断性,进而降低购买意愿。在理论方面,扩充了评论操控的研究。将默认好评纳入评论操控研究范畴,揭示了不同在线评论操控行为之间的差异。在对评论操控的作用机制上,从默认好评的特性出发,将感知产品诊断性作为解释机制应用于默认好评效应的解释中,为评论操控的后续研究提供了理论依据。从实践方面看,研究结果对网络购物平台评论体系的设计和管理有重要的借鉴意义。In online shopping scenarios,online review manipulation has increasingly become a prevalent phenomenon.Due to the lack of online reviews,many e-commerce platforms adopt default good review to manipulate online review.Unlike traditional online review manipulations,default good review has such characteristics as being uncontrollable by sellers and easily identifiable by consumers.Extant research on online review manipulation has not yet paid attention to the impact and mechanism of this special review manipulation.The study uses the real data of online product reviews and sales from one domestic e-commerce platform.Based on this panel data,we build a two-way fixed effect model,and explore the impact of the number of default good reviews on consumers’purchase behavior.Furthermore,this study discusses potential boundary conditions from the perspectives of both products(i.e.,product type,product popularity)and sellers(business model).In addition,this study also combines experimental method to examine the mediation role of product diagnosticity with one-factor between-group design.We set four conditions by manipulating the number of default good reviews displayed on product information page.Two hundred and thirty-four students participate in the study.The above-mentioned research design could make use of high external validity of second-hand data analysis and causal inference advantage of experiment method.The results show that,first,the number of default good reviews is negatively related to product sales.And when the number of default good reviews reaches a certain value,consumers’product purchases significantly decline.Second,product type,product popularity and business model significantly moderate the impact of the number of default good reviews on sales.Specifically,with regard to search products(vs.experiential products),less popular products(vs.more popular products)and products sold by distributors(vs.suppliers),the number of default good reviews has greater negative impact on product sales.Third,consumers�

关 键 词：评论操控 默认好评 产品诊断性 购买行为 

分 类 号：F713.36[经济管理—产业经济]