Knowledge that Transforms

Charity crowdfunding campaigns rely on emotional appeals to elicit empathy-driven support, yet they often struggle to translate such appeals into fundraising success. Drawing on the elaboration likelihood model, this study examines how emotions conveyed through textual descriptions and facial images relate to fundraising outcomes in charity crowdfunding. Using data from a leading crowdfunding platform, we apply machine learning techniques to analyze emotional content across both textual and visual modalities. Our analyses indicate that emotions expressed in textual descriptions contribute more to predicting fundraising performance than emotions conveyed through facial images, consistent with the higher levels of cognitive elaboration typically associated with text processing. Notably, our results reveal that anticipation in text is negatively associated with the amount raised, whereas sadness conveyed through facial expressions is positively associated with fundraising outcomes. To interpret these patterns, we introduce affective modality alignment, emphasizing that emotional appeals tend to be more effective when aligned with the processing tendencies of their delivery modality. By integrating machine learning with the ELM framework, this study advances understanding of emotion-based persuasion in digital philanthropy. Practically, it provides guidance for designing charity crowdfunding campaigns in cultural settings influenced by collectivist norms and offers insights that extend to similar institutional and sociocultural contexts.

In online knowledge communities (OKCs), humans and bots collaborate at massive scale, yet how their distinct resources shape collaboration patterns remains insufficiently understood. In particular, it is unclear whether humans consistently direct collaborative processes or whether bots can assume leading roles. Grounded in coordination theory, this study applies gSpan and local process model mining to 200 Wikipedia articles to examine how flow, fit, and sharing dependencies shape human-leadoff and bot-leadoff collaboration patterns. The results show that bots can function as initiating actors, guiding task flows in ways comparable to humans. Fit and sharing dependencies are key pathways linking human and bot resources to collaboration patterns. Task complexity exerts a moderating influence, revealing boundary conditions for coordination in human–bot collaboration. This study extends coordination theory to open, human–bot collaboration contexts and provides actionable guidance for managing task dependencies and fostering effective collaboration between humans and bots in OKCs.

The study of online community participation often categorizes behaviors as active or passive. Because information systems scholars view online communities as fluid organizations, the active-passive heuristic often misses the richer nuances of participation behaviors. We develop the visibility-cost framework (VCF) of online community participation, integrating eight participation behaviors that are essential for sustaining committed membership and user activity in online discussion communities. Based on commitment theory, we conduct an empirical study of Reddit users and illustrate how three forms of commitment impact participation on subreddits. Results of the study identify influences of affective commitment on lower-cost behaviors, continuance commitment on lower-visibility and higher-cost behaviors, and normative commitment on higher-visibility and higher-cost behaviors. We also highlight the relationship between lower-cost and higher-cost behaviors at each visibility level. The VCF expands the ability to increase online community participation by enabling targeted interventions that match specific commitment types to desired participation behaviors.

The integration of artificial intelligence (AI) with deep-tech products has been a transformative force across various industries. The synergy between AI and innovation continues to drive the development of cutting-edge solutions that shape the future of technology and business. This study examines the firm’s strategic decision of whether to make its innovative AI product compatible with its rival’s incumbent product. A compatible strategy allows the firm to generate more sales of the deep-tech product. In contrast, adopting an incompatible strategy enables the firm to increase sales of its incumbent product by attracting consumers who seek compatibility with the firm’s deep-tech product. The compatibility decision, aimed at maximizing the firm’s overall profitability from both the deep-tech and incumbent products, hinges critically on factors such as the compatibility utility of using an incumbent product that is compatible with the deep-tech product, the marginal production cost of the deep-tech product, and network effects exhibited by the products. The firm is inclined to adopt the compatible strategy when the deep-tech product has a low marginal production cost and the compatibility utility is also low. When network effects are weaker than the misfit cost, it promotes a compatible strategy. Stronger network effects than the misfit cost lead to one of the two firms monopolizing the incumbent market when consumers are rational, eliminating the need for a compatibility decision. However, if consumers’ rationality is bounded, both firms coexist in the incumbent product market, and the compatibility decision depends on factors influencing consumers’ expectations of the incumbent products’ demands.

Online games are the most widely used information system application in the world, with more than 3 billion users. There are many game genres, each of which strives to provide compelling gameplay experiences by offering users a variety of affordances. As a result, understanding how each affordance influences gameplay routines and game loyalty in different genres is crucial for building theories about online games. We developed a theoretical framework of 15 motivational affordances that lead to four gameplay routines and tested it using games from four major game genres. Our survey of 1,075 users found that some affordances and routines are important for some genres but unimportant for others, and may even have unintended negative effects in some cases. This shows that in multi-user systems such as online games, affordances that enable users to engage in certain routines may increase their loyalty but reduce the loyalty of other users who are affected by those actions. Our research offers contributions for researchers developing theories for online games by enabling them to focus on the effective affordances within each genre and by identifying which affordances have generally positive or generally unintended negative effects for each genre. From a practical perspective, by showing which affordances to offer and which to avoid in each of the four genres, our research can guide practitioners in making more effective design decisions to attract and retain users. The findings may also have implications for the design of other multi-user systems and workplace gamified systems by showing which affordances are impactful for most users.

Integrating artificial intelligence (AI) into organizations introduces challenges, particularly when AI-generated recommendations conflict with human judgment. Drawing on the theory of cooperation and competition and cognitive dissonance theory, we examine how human-AI conflict (low vs. high) and human-AI goal interdependence (cooperation vs. competition) shape cognitive dissonance and AI aversion across critical and non-critical decisions. Using an experiment with 432 employees, results show that higher human-AI conflict increases cognitive dissonance, which in turn reduces epistemic and social motivation and heightens AI aversion. The impact of human-AI conflict on cognitive dissonance intensifies when AI is perceived as a competitor and when decisions are critical. Moreover, in critical decisions, perceiving AI as a competitor amplifies the dissonance arising from human-AI conflict, whereas in non-critical decisions, cooperative and competitive perceptions of AI produce similar dissonance levels. These findings underscore the importance of aligning AI deployment with users’ relational perceptions and decision criticality.

The number of mobile app startups is increasing, but these startups face an increasingly hypercompetitive environment. To ensure long-term returns on their investments, these startups need to mitigate new entry threats. This entails limiting competitors’ ability to imitate products. However, a minimal amount is known about the factors that influence competitors’ identification of imitation targets and their willingness to imitate. We propose a research model that focuses on these two aspects to understand new entry threats. We adopt the cognitive view of Signaling Theory to examine the impact of a startup’s signaling on new entry threats. Startups rely on various rhetorical and substantive signals to establish their high quality in the market. We argue that competitors are influenced by those signals when “carrying out” new entry threats. In our model, we incorporate the competitor’s cognitive perspective by hypothesizing about the influence of the competitor’s funding status and experience. We test the model using a dataset of 3,286 startups active on the two largest app marketplaces from 2012 to 2022. The findings support all the hypotheses proposed. Thus, this study offers significant theoretical contributions and practical implications for mitigating new entry threats.

Insufficient patient disclosure and persistent stigma undermine effective mental health care; a challenge magnified during the COVID-19 pandemic. Telehealth offers a promising avenue to reduce access barriers and improve equity, yet its effectiveness depends on patients’ willingness to disclose sensitive information online. This study develops a middle-range, contextually adapted version of the disclosure processes model (DPM) to explain and predict how stigma and technological features shape online self-disclosure in mental health settings. We conducted a randomized web-based experiment with 309 participants who viewed a video vignette depicting a consultation between a patient and a psychiatrist. The vignette manipulated diagnosis (attention-deficit/hyperactivity disorder [ADHD] vs. schizophrenia) and consultation mode (in-person vs. virtual). Results show that willingness to disclose increases with greater trust in technology, higher perceived social presence, and richer communication media. Initial disclosure goals align with differing levels of technological trust and self-disclosure. However, perceived stigma weakens these positive relationships, reducing patients’ readiness to share sensitive information. The research advances theory by extending the DPM into a context-specific, middle-range information systems framework that integrates stigma and media characteristics in online mental health care. Practically, the findings identify key communication features—such as social presence, richness, and trust in telehealth platforms—that can be calibrated to foster disclosure of stigmatized information. These insights inform the design and implementation of telehealth services that promote open communication and improve treatment engagement in mental health and other stigma-laden domains.

Most user-generated content platforms utilize peer evaluation systems that incorporate upvotes and downvotes to distinguish high-quality from low-quality content. However, downvotes are often misused to target others and discourage content creators, and some platforms, such as Amazon and TripAdvisor, have removed the downvote option altogether. Meanwhile, the real-world impact of this platform-level intervention on user contributions remains underexplored, and we fill this gap by analyzing a restaurant review platform that disabled downvotes. We apply the regression discontinuity in time (RDiT) method and find that removing downvotes reduces the quantity of reviews but increases their quality. Further mechanism analysis suggests that the former may stem from perceived unfairness, and the latter from a motivation of continuing reviewers to stand out. Additionally, our heterogeneity analysis shows that the decline in review quantity is stronger among long-tenure reviewers, whereas the improvements in multiple quality dimensions are concentrated among reviewers with fewer followers. Our study reveals the nuanced behavioral effects of a platform-level design change and offers practical insights for designing peer evaluation systems that balance fairness, motivation, and content value.

Venture capital (VC) plays a key role in enabling information technology (IT) startups, driving economic growth and innovation. However, two defining features in the IT industry—deregulated market dynamics and gender-imbalanced labor structures—change VC effectiveness. The male-dominated ecosystem channels resources toward male entrepreneurs, leaving female-led IT startups less resourced and, in theory, more likely to benefit from VC support. Yet, they face challenges converting VC into internal capacity, such as hiring in male-dominated markets—a dynamic we term VC utilization bias. Deregulation further complicates this relationship: while it can enhance resource flexibility, it also intensifies competition, amplifying disparities. Using multiple datasets and the Telecommunications Act of 1996 as an exogenous shock, we document VC utilization bias and show that deregulation magnifies it. This work contributes to the information systems (IS) literature by examining gender bias at the executive level in the IT industry and how VC support amplifies gender disparity among IT entrepreneurs.