JHJian-Ren Hou

Jian-Ren Hou

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Jian-Ren Hou has a total of 11 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Language-Embedded Cultural Norms in LLM Economic Decision-Making Behavior

    by Xianghua (Jason) Wu, Kay-Yut Chen, Diana Wu, Jie (Jennifer) Zhang, Jian-Ren Hou

    Large Language Models (LLMs) are increasingly deployed across diverse linguistic and cultural contexts, yet it remains unclear whether cultural values embedded in language influence how these systems behave in social interactions. This paper examines how language and culture shape LLM decision-making in the context of economic negotiations. We prompt agents instantiated from multiple versions of LLMs to play a modified ultimatum game. In this game, one agent proposes how to divide resources, while the other specifies the minimum acceptable offer; the responder’s veto power is systematically varied across experimental conditions. Prompts are presented in 17 natural languages representing diverse cultural backgrounds. We report three main findings. First, agents’ offers vary significantly across languages, suggesting that the linguistic context of a prompt influences fairness-related behavior. Second, consistent with patterns observed in human participants, agents make higher offers and set higher acceptance thresholds when responders possess greater veto power. Third, both offers and acceptance thresholds are significantly associated with the Power Distance Index (PDI) of the cultures linked to each language, connecting LLM behavior to real-world cultural attitudes toward hierarchy and inequality. The positive relationship between Power Distance and offer size is consistently observed across all evaluated models. For acceptance thresholds, the overall association remains significant, although the strength of the effect varies across model families. Together, these findings provide evidence that culturally embedded biases are reflected in LLM behavior and may affect fairness in multilingual and cross-cultural applications. We discuss implications for the development and deployment of AI systems that operate equitably across linguistic and cultural settings.

  • Open Access

    Emergent Social Bias in AI and Mitigation Strategy

    by Diana Wu, Kay-Yut Chen, Xianghua (Jason) Wu, Jian-Ren Hou, Jie (Jennifer) Zhang

    Bias in artificial intelligence is commonly viewed as a reflection of human and societal prejudices embedded in training data. This study instead examines whether bias can emerge endogenously from interaction and learning among AI agents, even in the absence of human-generated data or societal input. Using deep reinforcement learning (DRL) agents trained solely through repeated interactions with one another in a trust game, we find that the agents autonomously develop in-group favoritism and out-group discrimination, which reduces overall trust and trustworthiness at the population level. We evaluate multiple mitigation strategies and find that training-based interventions, such as inclusive training and retraining with unbiased AI coaches, reduce group disparities but fail to improve, and often worsen, cooperation among AI agents. In contrast, an architectural intervention that disrupts internal neural representations can eliminate group biases while maintaining high levels of cooperation among agents. To understand the origin of emergent bias in AI, we introduce an information-theoretic framework based on mutual information. Our analysis suggests that group identity information and behaviorally relevant information may be encoded differently in DRL agents' neural networks. These findings motivate new design principles for building socially intelligent and cooperative multi-agent AI systems.


  • The service attributes of robo-advisors: a choice-based conjoint analysis

    by Jian-Ren Hou, Yen-Hsi Li, Sarawut Kankham

    Purpose: As an alternative to hiring financial specialists or investment consultants, robo-advisors offer financially automated investment services. This study aims to investigate how robo-advisors' service attributes, risk attitude and financial self-efficacy influence customers' choice preferences of adopting robo-advisors.Design/methodology/approach: Two hundred fifty-one online surveys were used to collect data, and choice-based conjoint analysis was conducted.Findings: Results show that increasing annual fees negatively impact customers' choice preferences. Promotion, general investment education and additional human assistance have a positive impact. Furthermore, risk-seeking and risk-averse customers require more human assistance than risk-neutral customer and customers with high levels of financial self-efficacy prefer more general investment education and additional human assistance than those with lower levels. In addition, customers in the older age group prefer promotion, general investment education and additional human assistance, while wealthy customers prefer lower annual fees, higher general investment education and more additional human assistance compared to middle-class and low-income groups.Originality/value: This study contributes to robo-advisor providers to provide appropriate service attributes for each customer group.


  • Follow and spread the word: the effects of avatars and message framing in promoting fact checking posts on social media

    by Jian-Ren Hou, Sarawut Kankham

    Purpose: Fact-checking is a process of seeking and displaying facts to confirm or counter uncertain information, which reduces the spread of fake news. However, little is known about how to promote fact-checking posts to online users on social media. Through uncertainty reduction theory and message framing, this first study examines the effect of fact-checking posts on social media with an avatar on online users' trust, attitudes, and behavioral intentions. The authors further investigate the congruency effects between promotional message framing (gain/loss/neutral) and facial expressions of the avatar (happy/angry/neutral) on online users' trust, attitudes, and behavioral intentions in the second study.Design/methodology/approach: The authors conducted two studies and statistically analyzed 120 samples (study 1) and 519 samples (study 2) from Facebook users.Findings: Results showed that including the neutral facial expression avatar in fact-checking posts leads to online users' greater trust and more positive attitudes. Furthermore, the congruency effects between loss message framing and the angry facial expression of the avatar can effectively promote online users' trust and attitudes as well as stronger intentions to follow and share.Originality/value: This study offers theoretical implications for fact-checking studies, and practical implications for online fact-checkers to apply these findings to design effective fact-checking posts and spread the veracity of information on social media.


  • More than feelings? How Facebook reaction icons affect online users' behavioral intentions toward online health rumor posts

    by Jian-Ren Hou, Sarawut Kankham

    When the spread of online health rumors on social media causes public concerns, the public is calling for action. However, little study has investigated how Facebook reaction icons (expressing feelings function) affect online users' behavioral intentions (intention to trust and share) toward online health rumor posts. The current study addresses this gap by focusing on the effect of Facebook reaction icons in two conditions: Facebook reaction icons' presence (versus absence), and Facebook reaction icons' emotional valence (positive versus negative versus neutral). Moreover, the authors also investigated the interaction between Facebook reaction icons' emotional valence and online health rumor posts' framing headlines (gain versus loss). The authors used a 7 (Facebook reaction icons: Love, Like, Haha, Wow, Sad, Angry, and no icon) × 2 (Facebook framing headlines: gain and loss) between-subjects design, analyzing 507 samples from online users with one-way ANOVA and MANOVA. Results show that online health rumor posts without Facebook reaction icons are more likely to negatively change online users' behavioral intentions than the posts with Facebook reaction icons; negative reaction icons (Sad and Angry) lower online users' behavioral intentions than positive reaction icons (Love and Like). Further, the incongruency effect of interaction (i.e. positive reaction icons with a negative message) would have more negative effects on online users' behavioral intentions than the congruency effect (i.e. positive reaction icons with a positive message). This study has rich contributions to theoretical and practical implications for the Facebook platform and Facebook users to apply Facebook reaction icons against online health rumor posts.


  • The advertising communication effectiveness of using netizen language code-switching in Facebook ads

    by Ying-Feng Kuo, Jian-Ren Hou, Yun-Hsi Hsieh

    Purpose: Netizens refer to citizens of the internet, and code-switching refers to the use of more than one language, style or form of expression to communicate. This study explores the advertising communication effectiveness of using netizen language code-switching in Facebook ads. Moreover, if a brand is with negative brand images, using positive brand images as a control group, this study investigates not only the advertising communication effectiveness of netizen language code-switching but also its effectiveness of remedying the negative brand images.Design/methodology/approach: Online experiments were conducted, and data were analyzed using independent sample t-test, MANOVA and ANOVA.Findings: The results indicate that netizen language code-switching can enhance advertising communication effectiveness in Facebook ads. Furthermore, under a negative brand image, netizen language code-switching has significant effects on improving Facebook advertising communication effectiveness.Originality/value: This study takes netizens as the research subjects to explore the advertising communication effectiveness of netizen language code-switching in Facebook ads. This study provides further insight into the effect of netizens' culture on Facebook advertising and enriches the existing literature on social media advertising, as well as expanding the application of code-switching. The results of this study provide enterprises a new perspective on the copywriting content design of Facebook ads.


  • The effects of anchoring on backers' pledge in reward-based crowdfunding: evidence from Taiwanese market

    by Ying-Feng Kuo, Cheng-Han Lin, Jian-Ren Hou

    Purpose: Crowdfunding allows enterprises or individuals to collect funds from numerous other individuals. This study applies the anchoring effect and range theory in reward-based crowdfunding to explore how different pledge option designs affect the backers' final pledge amount. Moreover, this study examines whether showing the current average amount pledged in the fundraising process has an anchoring effect on the subsequent backers' pledge amount.Design/methodology/approach: Online experiments were conducted, and data were analyzed using the Kruskal–Wallis test and Spearman rank correlation analysis.Findings: Results show that among the three pledge option designs, employing the “bolstering range offer” has the highest backing amount. However, presenting the current average amount pledged in the fundraising process has a reversed anchoring effect on subsequent backers' pledge amount only in the case of a crowdfunding project in the physical goods category with a “point offer.”Originality/value: To the best of authors’ knowledge, no reward-based crowdfunding platform has yet provided the pledge option design of a “bolstering range offer.” This study reveals that the “bolstering range offer” can significantly increase the amount pledged. This study extends the crowdfunding research area to crowdfunding success and suggests a novel way to set up pledges.