Date of Award
2025
Degree Type
Thesis
Degree Name
Doctor of Business Administration (DBA)
Abstract
线索是消费者在线购买决策的基础，在线评论是一个外在线索，是消费者发布的购买或使用体验的反馈，是减少在线购物信息不对称的有效方式。研究在线评论对消费者购买决策的影响机制具有重要的理论意义和商业价值。
本研究以北京高校的31名大学生为被试，以6类商品图片为实验材料，以框架水平、评价比例和评级高低为自变量，以消费者的购买意愿及ERP成分的振幅为因变量，采用2×2×2重复测量实验设计，结合行为和脑电技术，探讨了在线评论对消费者购买决策影响的行为和神经机制。
行为实验结果显示：高从众框架下，评价比例与评级高低的交互效应显著。即在五星评级时，好评率99%与差评率1%商品的购买意愿差异显著；而一星评级时，二者差异不显著。低从众框架下，评价比例与评级高低的交互效应显著。即五星评级时，好评率90%与差评率10%商品的购买意愿差异显著；而一星评级时，二者不但差异显著，而且差异效应更大。
行为实验结果表明：当在线评论中消费者意见一致性高时，评价比例在高评级时是一个有效的购买决策利用线索，在低评级时则线索功能失效；而当在线评论中消费者意见一致性低时，无论评级高低，评价比例都是一个有效的购买决策利用线索。
P2分析显示：高从众框架下，好评率99%与差评率1%商品的P2振幅差异显著，差评率1%的P2振幅更大；在低从众框架下，五星与一星评级商品的P2振幅差异显著，一星评级的P2振幅更大。N400分析显示：高从众框架下，好评率为99%且一星评级商品的N400振幅最高；低从众框架下，好评率为90%且一星评级商品的N400振幅最高。LPP分析显示：高从众框架下，好评率99%与差评率1%的商品差LPP振幅异显著，差评率1%的LPP振幅更大；但在两种框架下，LPP成分都显著区分了五星与一星评级商品，五星评级的LPP振幅更大。
ERP实验结果表明：消费者在线购买决策中，P2成分可以表征“风险感知”，N400成分可以表征“风险重估”，LPP成分可以表征“满意分类”。基于在线评论的消费者购买决策是一个“风险感知”、“风险重估”和“满意分类”的S-O-R信息加工过程。
Cues are the foundation of consumers' online purchasing decisions. Online reviews serve as external cues, providing feedback on consumers' purchase or usage experiences and effectively reducing information asymmetry in online shopping. Studying the impact mechanism of online reviews on consumer purchasing decisions has significant theoretical and commercial value.
This study used 31 college students from a Beijing university as participants, with six types of product images as experimental materials. The independent variables were the level of framing, the ratio of positive reviews, and the rating levels. The dependent variables were consumers' purchase intentions and the amplitude of ERP components. A 2×2×2 repeated-measures experimental design was adopted, combining behavioral and electroencephalographic techniques to explore the behavioral and neural mechanisms underlying the influence of online reviews on consumer purchasing decisions.
The behavioral experiment results showed that under high consensus framing, the interaction effect between the ratio of positive reviews and rating levels was significant. Specifically, when the rating was five stars, the difference in purchase intention between products with a 99% positive review rate and those with a 1% positive review rate was significant; however, when the rating was one star, the difference was not significant. Under low consensus framing, the interaction effect between the ratio of positive reviews and rating levels was also significant. Specifically, when the rating was five stars, the difference in purchase intention between products with a 90% positive review rate and those with a 10% positive review rate was significant; when the rating was one star, the difference was not only significant but also larger.
The behavioral experiment results indicated that when the consistency of opinions in online reviews was high, the ratio of positive reviews was an effective cue for purchasing decisions when the rating was high; when the rating was low, the cue function failed. When the consistency of opinions in online reviews was low, regardless of the rating level, the ratio of positive reviews was an effective cue for purchasing decisions.
ERP analysis showed that under high consensus framing, the P2 amplitude difference between products with a 99% positive review rate and those with a 1% positive review rate was significant, with the P2 amplitude being larger for products with a 1% positive review rate. Under low consensus framing, the P2 amplitude difference between five-star and one-star rated products was significant, with the P2 amplitude being larger for one-star rated products.
N400 analysis showed that under high consensus framing, the N400 amplitude was highest for products with a 99% positive review rate and one-star ratings; under low consensus framing, the N400 amplitude was highest for products with a 90% positive review rate and one-star ratings.
LPP analysis showed that under high consensus framing, the LPP amplitude difference between products with a 99% positive review rate and those with a 1% positive review rate was significant, with the LPP amplitude being larger for products with a 1% positive review rate. In both framing conditions, the LPP component significantly distinguished between five-star and one-star rated products, with the LPP amplitude being larger for five-star rated products.
ERP experimental results indicated that in online purchasing decisions, the P2 component can represent "risk perception," the N400 component can represent "risk re-evaluation," and the LPP component can represent "satisfaction classification." Based on online reviews, consumers' purchasing decisions involve a process of "risk perception," "risk re-evaluation," and "satisfaction classification" in the S-O-R information processing model.
Keywords
在线评论, 事件相关电位, 消费者购买意愿, 线索利用理论, 前景理论, Online reviews, Event-related potentials, Consumer purchase intention, Cue utilization, Prospect theory
Language
Chinese (Simplified)
Copyright
The copyright of this thesis is owned by its author. Any reproduction, adaptation, distribution or dissemination of this thesis without express authorization is strictly prohibited.
Recommended Citation
王永平 (2025)。在线评论中，从众框架与星级评分如何影响消费者购买决策 : 行为与事件相关电位分析 (博士論文，香港嶺南大學)。檢自 https://commons.ln.edu.hk/otd_tpg/69/
Comments
How conformity framing and star ratings in online reviews influence consumer purchase decisions : a behavioral and event-related potential analysis