Research On The Influence Of Different Recommendation Methods On Consumer Purchase Decision-making Process

The rapid growth of e-commerce makes online shopping an important way of consumption.An e-commerce platform can use a recommendation algorithm to provide consumers with high-quality services that meet their personal preferences and personalized needs so that consumers can find their favorite products from the complex commodity information,so as to save the cost of commodity search and improve the efficiency of commodity selection.As an effective way to solve the information overload in the Internet era,the recommendation algorithms have become a hot topic in academia and industry and has a profound impact on the development of the online shopping market.Although the recommendation algorithms bring considerable technology bonus to the online shopping market,some researchers found that people have resistance to the algorithm recommendations.Therefore,researchers began to pay attention to whether people trust algorithm recommendations or human recommendations.The existing comparative studies of algorithm recommendations and human recommendations have inconsistent research conclusions.Some studies found that algorithm recommendations are superior to human recommendations in prediction accuracy and scalability,but people do not trust algorithm recommendations.Some studies also show that people trust algorithm recommendations more and tend to rely on algorithm recommendations for decision-making.In order to solve these differences and controversies,researchers in the fields of economics and management have been using questionnaires and behavioral experiments to study the influence of algorithm recommendations and human recommendations on individual decision-making behavior.However,due to the fact that the "black box" behind individual behavior cannot be involved and the lack of research on consumer decision-making behavior,this topic is still a science that needs to be further explored and studied in the consumer decision-making research problem.Therefore,the academic and industrial areas explore new research methods and use the theories,tools,and methods of neuroscience to solve these problems.The theories and tools of neuroscience can deepen the research of consumer decision-making from behavior level to the neuropsychological process level so that we can have a deeper insight into the deep reasons behind consumer purchase decision-making.The purpose of this paper is to explore the neural mechanism of the influence of different recommendation methods on consumers’ purchase decision-making process by using eye-tracking technology and electroencephalography technology.Eye-tracking technology and electroencephalography technology,as the mainstream technology of neuroscience research,have the characteristics of dynamic continuous data recording,high time accuracy,wide intuition,and strong objectivity.This study constructs the model of psychological factors in the process of consumer’s purchase decision under the influence of different recommendation methods based on the theory of consumer purchase decision-making model,the five-stage consumer decision-making process and Stimulus-Organism-Response,and divides the consumer purchase decision-making process into three stages: information search,evaluation,and purchase decision.In terms of eye-tracking and electroencephalography corresponding neuroscience indicators,the internal psychological factors of consumers in the three stages are objectively measured,and the differences of consumers’ preferences for different recommendation methods and the underlying psychological mechanism are revealed from the perspective of the neuropsychological process.This paper focuses on three studies: one is to explore the neural mechanism of different recommendation methods influencing consumer information search behavior around the information search stage;the other is to explore the neural mechanism of different recommendation methods influencing consumer evaluation behavior around the evaluation stage;the third is to explore the neural mechanism different recommendation methods influencing consumer purchase decision behavior around the purchase decision stage mechanism.The main findings of this paper are as follows:(1)In the information search stage,the algorithm recommendations can improve consumers’ attention,cognitive integration,working memory capacity,emotional arousal and trust to commodity information.The results of the neuropsychological process show that the algorithm recommendations make consumers pay more attention to commodity information,which is manifested in shorter first fixation time and higher theta frequency band activity in the occipital region of the scalp;the algorithm recommendations make consumers have higher cognitive integration to commodity information,which is manifested in higher beta frequency band activity in the temporal region of the scalp;the algorithm recommendations make consumers have higher working memory capacity for commodity information,which is manifested in shorter average fixation time and higher gamma frequency band activity in the parietal region of the scalp;the algorithm recommendations make consumers have higher emotional arousal degree for commodity information,which is manifested in larger average left eye pupil size and higher beta frequency band activity in the frontal region of the scalp;the algorithm recommendations make consumers have better performance in the trust degree of commodity information,which is manifested in higher beta frequency band activity in the prefrontal region of the scalp.(2)In the evaluation stage,the algorithm recommendations can improve the working memory capacity,emotional arousal,motivation,and trust of consumers when they perform click behavior.The results of the neuropsychological process show that the algorithm recommendations make consumers have higher working memory capacity when they perform click behavior,which is manifested in higher gamma band activity in the parietal region of the scalp;the algorithm recommendations make consumers have higher emotional arousal degree when they perform click behavior,which is manifested in larger average left eye pupil size and higher beta frequency band activity in the frontal region of the scalp;the algorithm recommendations make consumers have a higher motivation to perform click behavior,which is manifested in higher beta frequency band activity in the left frontal region of the scalp;the algorithm recommendations make consumers have a higher trust when performing click behavior,which is manifested in higher beta frequency band activity in the prefrontal region of the scalp.(3)In the purchase decision stage,the algorithm recommendations can improve the working memory capacity,emotional arousal,motivation and trust of consumers.The results of the neuropsychological process show that the algorithm recommendations make consumers have higher working memory capacity when making the purchase decision,which is manifested in shorter average fixation time;the algorithm recommendations make consumers have higher emotional arousal degree when making the purchase decision,which is manifested in larger average left eye pupil size and higher beta frequency band activity in the frontal region of the scalp;the algorithm recommendations make consumers have higher motivation when making the purchase decision,which is manifested in higher beta frequency band activity in the left frontal region of the scalp;the algorithm recommendations make consumers have higher trust when making the purchase decision,which is manifested in higher beta frequency band activity in the prefrontal region of the scalp.(4)Consumers have an "algorithm recommendations trust effect" on algorithm recommendations.The results of the three studies suggest that,compared with human recommendations,the algorithm recommendations can promote consumers’ information search behavior,evaluation behavior,and purchase decision-making behavior.According to the above results,this paper draws the following conclusions: consumers have an "algorithm recommendations trust effect" on the algorithm recommendations,that is,compared with the human recommendations,consumers trust algorithm recommendations more,and they tend to rely on algorithm recommendations to make purchase decisions.