Files in this item



application/pdfTIAN-DISSERTATION-2020.pdf (5MB)Restricted to U of Illinois
(no description provided)PDF


Title:Cross-cultural recommender system use and design
Author(s):Tian, Kathy
Director of Research:Yao, Mike
Doctoral Committee Chair(s):Yao, Mike
Doctoral Committee Member(s):Nelson, Michelle; Belk, Russell; Ham, Chang-Dae
Department / Program:Inst of Communications Rsch
Discipline:Communications and Media
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Cross-Culture, Recommendation Systems, User Design
Abstract:Recommendation engines have become a pervasive component of consumers’ everyday life. These recommender systems are employed on e-commerce sites across the world. However, despite their pervasive presence, the user interface design, and the way in which products are recommended, these systems are relatively similar across cultures. Across three studies, I explored how culturally informed cues, both social and visual, affect preferences for a recommender systems’ style and design. Specifically, I examined how values of individualism-collectivism and differences in information processing influence consumer behavior in the recommendation system used from both within and between-cultures perspective (e.g., individual-level and national culture). Consumers hailing from collectivistic cultures (e.g., China) and those who exhibit greater collectivistic qualities at the individual level are believed to prioritize the needs of the ingroup over individual preferences compared to their individualistic counterparts. Collectivists have been found to be more concerned with peer perceptions and more susceptible to social influence on social networking sites compared to individualists. To understand whether these values would affect consumers in the context of e-commerce recommender systems, participants were exposed to two recommender system design conditions; they viewed products that were either recommended by a socially similar character or by a generic recommendation engine. Similarly, individuals hailing from East Asia are believed to attend to and process information more holistically and incorporate more pieces of information relative to Westerners. These differences in processing styles stem from tendencies toward holistic or analytical information processing, where holistic processing involves attaching relationships between objects, and analytical processing involves focusing categorically on focal objects. To investigate how information processing influences consumers, both within and between-cultures, participants viewed either a recommender with high information density or low information density. The results of the studies showed that there were no interaction effects between the level of individualism-collectivism, information processing, or national culture and condition. Individuals scoring higher for collectivism and Chinese (vs. American) participants did not prefer a recommendation system where a socially similar character recommended the products. Likewise, Chinese (vs. American) participants and those scoring higher for holistic tendencies did not display a preference for a more information-dense recommender. These findings were consistent using both self-reported responses and when behaviors were tested on a website shopping platform. Additionally, Chinese and American participants did not consistently differ in their level of individualism, which may be a function of China’s socio-economic landscape. Taken together, the results of three-part studies suggest that cultural information cues may not be as salient in online settings. This research contributes to the on-going debate about whether Hofstede’s cultural dimensions and cultural differences are stable over time. It also addresses questions of how new technologies influence culture and behaviors.
Issue Date:2020-05-06
Rights Information:Copyright 2020 Kathy Tian
Date Available in IDEALS:2020-08-26
Date Deposited:2020-05

This item appears in the following Collection(s)

Item Statistics