Knowledge that Transforms

OES

Despite the growing popularity of online referral programs, a minimal amount is known regarding the theoretical foundations that drive the key actions associated with successful referrals. In this paper, we study which type of referral reward structure is most effective in maximizing word-of-mouth by conducting two randomized experiments in mobile gaming context. Specifically, we examine the effect of three incentive schemes: selfish reward (inviter gets all the reward), equal-split reward (50-50 split), and generous reward (invitee gets all the reward). Consistent across the two experiments, we find that pro-social referral incentive schemes, namely the equal-split and generous schemes, tend to dominate purely selfish schemes in creating WOM. Our mechanism-level analysis shows that both equal-split and generous schemes result in higher number of conversions by significantly increasing the invitee’s likelihood to accept referrals, which we further show that is partially due to selective and better targeted referrals. Our results contribute to the understanding of the optimal design of online referral programs and provide important implications for designing effective referral reward schemes in the digital world.

The sharing economy has fundamentally changed the way many individuals work. In this paper, we study the impact of the entry of a major ridesharing platform into U.S. Metropolitan Statistical Areas (MSAs), on the supply and demand sides of the labor market. Leveraging the difference-in-differences (DID) research design and a data set combining multiple U.S. Census archival sources, we exploit the variation in labor market metrics before and after Uber’s entry into the MSAs. Our empirical findings reveal that the introduction of the ridesharing platform has an empowering effect on workers (the supply side of the labor market) and a competition effect on traditional jobs (the demand side of the labor market). Specifically, Uber’s entry into the MSAs increases labor force participation, decreases the unemployment rate of residents living below the poverty level, and improves the employment and financial status of low-income workers. In addition, Uber’s entry reduces the employment number and increases wages of conventional low-skill and/or low-wage jobs. This paper provides empirical evidence of the impact of a digital sharing economy platform on the labor market and suggests that policymakers and platform operators should account for this broader impact when they devise policies and make strategic decisions.

Inadequate patient safety is a serious issue in current medical practice. Medical errors cause adverse events (AEs) for patients and lead to premature deaths, unintended complications, prolonged hospital stays, and higher medical costs. Although the importance of AE prediction and prevention is well recognized in the information systems literature, there is a dearth of research on modeling and predicting AEs caused by medical errors. Following the design science research paradigm, this study describes the search, design, and evaluation of a novel in-hospital AE prediction model, called Stochastic Autoregressions for Latent Trajectories (SALT). The proposed model uniquely integrates generalized linear mixed model with multitask learning and stochastic time-series processes. Results from our empirical evaluation show that SALT outperforms prior state-of-the-art techniques in predicting AEs during patients’ hospital stays. Through a simulation, we further demonstrate significant cost savings potential when hospitals implement and integrate SALT in their inpatient care. This study contributes to the design science literature by formalizing the in-hospital AE prediction problem, on the one hand, and developing a novel graphical model to address the prediction problem, on the other. For healthcare practitioners and administrators, our predictive analytics approach unveils important insights to minimize AEs.

Blockchain is an emerging technology that enables two or more entities to conduct secure transactions. After a blockchain transaction is executed, it cannot be altered because the transaction infor...

Ensuring organizational regulatory and legal compliance is a challenging, high-stakes management task. Overlooking or underestimating noncompliance in organizations can result in substantial fines, damage to a company’s reputation, and, ultimately, loss of business. To help alleviate this risk, we explore whether organizations can assess noncompliance by monitoring users’ answers and related mouse cursor movements in an intelligent online questionnaire. Namely, we propose that noncompliant (compared with compliant) individuals will experience more cognitive dissonance on questions about (1) what constitutes noncompliance behavior and (2) what consequences for noncompliance are appropriate. We predict that increased cognitive dissonance in noncompliant individuals will influence their questionnaire responses—as well as their mouse cursor movements—when answering these questions. We collect data to test our hypotheses in a study from individuals who voluntarily chose to cheat on an online task for monetary gain. The results suggest that the responses to these two groups of questions, along with the associated mouse cursor movement, can work together as a low-cost, scalable tool to help assess noncompliance risk. This paper contributes to theory by explaining how people who are noncompliant tend to provide more lenient answers to questions about what constitutes compliance and the consequences of noncompliance. In addition, we show how mouse-cursor deviation provides theoretical insight into the level of cognitive dissonance that users experience, and that users who are noncompliant show greater deviation on compliance and consequence questions.

The means by which e-commerce websites can reach and track online customers have expanded enormously through the use of various digital marketing referral channels. However, evaluating comparative effectiveness and return on investment (ROI) across different referral channels remain difficult undertakings for many companies. This study aims to contribute to this line of investigation by quantifying the relative effectiveness, the dynamics, and the interdependencies among three types of major online referral channels: search engines, social media, and third-party websites. To this end, we employ the vector autoregressive (VARX) model on a large-scale clickstream dataset and have the following findings. Though search engine referrals demonstrate strong impact on sales, our results show that social media referrals have the strongest immediate and cumulative effects on e-commerce websites’ conversion rates. Our results also demonstrate the synergies and interdependencies across these channels. This study contributes to the multi-channel analytics literature and sheds new light to digital marketing managers on assessing the cumulative impact and the economic value of online referral channels.

A reliable and accurate estimate of the expected hospital length of stay (LOS) of a patient is important to patients, medical providers, and insurance companies. Predicting hospital Length of Stay (LOS) is a complex and ill-structured problem, driven by many factors such as a patient’s individual characteristics, treatment plans, and disease-interactions. In this paper, we develop a novel model to predict the expected LOS at the time of admission by combining network science and deep learning. We propose a two-dimensional construct of latent comorbidities comprising historical and probable comorbidities that a patient does not currently manifest but could likely develop during the course of hospital stay. The probable comorbidities are derived from a network comprising relationships among diseases observed in 3.2 million patient records in hundreds of US hospitals. We employ this construct of latent comorbidities in deep learning models to predict patients’ LOS using almost 10 million other patient visits belonging to various disease categories. Implementing these models and analyses required a high-performance computing (Big Data) facility. The average mean absolute percent error of our models across all categories of diseases was 29.8%, which is the best in the current state-of-the-art. Our primary contribution is in developing a generalizable method to create a predictor construct for recognizing underlying relationships through network analyses, which can then be used in a deep learning model to predict an exogenous dependent variable.

The online generation and dissemination of false information (e.g., through Facebook, Twitter, Snapchat and other Internet media), commonly referred to as “fake news”, has garnered immense public attention following the 2016 Brexit referendum, three US elections, the 2019 Indian lynchings, and the 2019 rise in polio cases in Pakistan. Fake news undermines public life across the globe, especially in countries where journalistic practices and institutions are weak [3]. Some fake news is created to spread ideological messages or to create mischief, whereas other fake news is created for profit, such as the Macedonian teenagers who created fake news sites during the 2016 US election to drive advertising [22]. Research shows that fake news spreads “significantly farther, faster, deeper, and more broadly” than true news [24:1146] and has had major societal impacts [15]. All signs indicate that it will get worse as political activists, scammers, alternative news media, and hostile governments become more sophisticated in their production and targeting of fake news. Fake news and other types of false information are also a matter of concern for business and management research and practice [2,9,10,11,12,19]. Businesses have engaged in deceptive communications such as greenwashing, astroturfing, false advertising and other types of false messages [4,5,14], but false content presented as news presents a novel range of issues for individuals, organizations, and societies [1,18]. The widespread adoption and use of information and communication technologies, particularly social and digital media, play a key role in the current wave of fake news and false information sweeping the globe [1,7,13]. We believe that the IS discipline can contribute significantly to the discourse, as it already has in related areas such as cyberdeviance [23,24] and deception [e.g., 5]. Our field can draw on its intellectual core of theories and empirical findings on the design, use, and impacts of IT artifacts at different levels of analysis. A nascent body of IS research on this topic is emerging [6,8,16–18,20,21]. Related areas such as review manipulation [e.g., 11] and social behaviors in online social networks [e.g., 10,12,21] can provide valuable lessons to apply to online fake news and false information more generally. Yet there is a dearth of evidence about many aspects, and many issues remain open to debate. We received 80 submissions, which went through three rounds of review and revision. The papers spanned a diverse set of experimental, qualitative, econometric, and analytical methods, and focused on fake news around the globe. The set of accepted papers are also diverse in methods and focus. We would like to say that collectively the articles offer several viable solutions to the problem of fake news. However, this is not the case in all instances.

As digital platform ecosystems grow in prominence, their interconnectedness and complexity also grow, making operational failure likely. How failures in such systems affect user perceptions of separate ecosystem components, however, is not well understood. This research investigates attribution of responsibility and discontinuance recommend ations for ecosystem components after failures of ambiguous origin. Building on attribution theory, platform ecosystems literature, and research on digital borders, we conducted two scenario-based experiments investigating negative consequences of failure for ecosystem components. We also explored contingent effects from design elements (border strength) and contextual factors (disruption severity). Results demonstrated that when failures occur, negative consequences diffuse to all ecosystem components, with apps receiving the strongest discontinuance recommendations. Greater disruption severity increased discontinuance recommendations for the app. Furthermore, border strength between ecosystem components shifted negative consequences for failure toward the platform (e.g., operating system [OS] and device). Perceptions of locus and controllability were the primary mechanisms driving attributions of responsibility for failure. However, contrary to attribution theory, lack of failure stability increased blame for the app instead of reducing it. Despite higher coordination costs, our results indicate the importance of better-integrated ecosystems that experience fewer faults and that app developers bear the greatest burden in delivering this experience. Furthermore, attribution for failure can be shaped by clearly delineated borders. Thus, design decisions affecting border strength should be actively managed by ecosystem participants, and app developers may be incentivized to elevate border strength.