What shall I learn? Two-stage decision making under social influence on corporate E-learning platforms

2023 | Decision Support Systems | Citations: 0

Authors: Song, Yiping Amy; Zhang, Lingling; Ma, Liye; Bose, Indranil

Abstract: E-learning platforms have increasingly been adopted by corporate employees in th ... Expand

Abstract: E-learning platforms have increasingly been adopted by corporate employees in the workplace. On these plat­ forms, users typically follow a two-stage decision-making process: they first choose which content to learn and then decide how much to continue learning. The decisions of individual employee users are influenced by members of the same workplace organization (group influence) and general users on the platform from other organizations (mass influence). Extant research has not shown how different types of social influence impact different decisions. Using data from a corporate e-learning platform, this study examines how group influence and mass influence support employees’ learning decisions, from the perspective of the elaboration likelihood model (ELM). The results reveal that mass users’ past choices only influence low-elaborative choice decisions but not high-elaborative engagement decisions. In contrast, workgroup members’ past choices influence both the low-elaborative choice and high-elaborative engagement decisions. Furthermore, positive synergy exists between the two types of social influence for the choice decision, but the synergy dissipates for the engagement decision. These findings can help online content platforms design appropriate information-sharing systems to influence users’ choice and engagement decisions. The results can also help corporates take advantage of social influence to motivate employees to engage in work-related online learning. Collapse

Topics: social influence online learning system decision making decision support system logistics management

Methods: descriptive statistic robustness check case control study synthetic control method paired t-test

Theories: elaboration likelihood model

Optimizing Real-Time Wellness Intervention Studies in Healthcare Shift Workers: Outcomes from a Prospective, Single-subject Design Study of a Digital Sleep Aid

2022 | HICSS | Citations: 0

Authors: Duggan, Nicole; Hasdianda, Mohammad Adrian; Jambaulikar, Guruprasad; Goldsmith, Andrew; Condella, Anna; Landry, Adaira; Boyer, Edward; Eyre, Andrew

Abstract: Sleep disorders associated with shift work impair alertness and increase risk of ... Expand

Abstract: Sleep disorders associated with shift work impair alertness and increase risk of chronic physical and mental health disease. Though studying sleep aids in atrisk populations such as shift workers is critical to improving wellness, implementing robust studies to evaluate wellness interventions in live populations can be particularly challenging. We performed a prospective, single-subject design study in which we assessed the effect of noise-masking earbuds on sleep quality, sleepiness, and stress level in health care shift workers. Despite being gifted the technology at the end of the study, we faced poor participant accrual and study retention. Additionally, robust analysis of our intervention’s impact on post-shift alertness was underpowered due to variable participant scheduling. In order to make meaningful advances in improving wellness among such groups, studies emphasizing objective outcome measures while minimizing participant obligation to study protocols are ideal. Here, we discuss our experience studying a digital sleep intervention in a real-time setting and propose solutions for improving subject accrual, retention and methods for objective data collection in studies performed in a live shift worker population. Collapse

Topics: health information system electronic mail

Methods: survey regression analysis method descriptive statistic experiment case control study

High-Performance Detection of Corneal Ulceration Using Image Classification with Convolutional Neural Networks

2021 | HICSS | Citations: 0

Authors: Gross, Jan; Breitenbach, Johannes; Baumgartl, Hermann; Buettner, Ricardo

Abstract: Corneal Ulcer, also known as keratitis, represents the most frequently appearing ... Expand

Abstract: Corneal Ulcer, also known as keratitis, represents the most frequently appearing symptom among corneal diseases, the second leading cause of ocular morbidity worldwide. Consequences such as irreversible eyesight damage or blindness require an innovative approach that enables a distinction to be made between patterns of different ulcer stages to lower the global burden of visual disability. This paper describes a Convolutional Neural Network-based image classification approach that allows the identification of different types of Corneal Ulcers based on fluorescein staining images. With a balanced accuracy of 92.73 percent, our results set a benchmark in distinguishing between general ulcer patterns. Our proposed method is robust against light reflections and allows automated extraction of meaningful features, manifesting a strong practical and theoretical relevance. By identifying Corneal Ulcers at an early stage, we aid reduction of aggravation by preventively applying and consequently tracking the efficacy of adapted medical treatment, which contributes to IT-based healthcare. Collapse

Topics: usability big data health information system computer hardware human computer interaction

Methods: machine learning descriptive statistic computational algorithm classification method case control study

Chronic Disease Management: How It and Analytics Create Healthcare Value Through the Temporal Displacement of Care

2020 | Management Information Systems Quarterly | Citations: 7

Authors: Thompson, Steve; Whitaker, Jonathan; Kohli, Rajiv; Jones, Craig

Abstract: The treatment of chronic diseases consumes 86% of U.S. healthcare costs. While h ... Expand

Abstract: The treatment of chronic diseases consumes 86% of U.S. healthcare costs. While healthcare organizations have traditionally focused on treating the complications of chronic diseases, advances in information technology (IT) and analytics can help clinicians and patients manage and slow the progression of chronic diseases to result in higher quality of life for patients and lower healthcare costs. We build on prior research to introduce the notion of temporal displacement of care (TDC), in which IT and analytics create healthcare value by displacing the time at which providers and patients make interventions to improve healthcare outcomes and reduce costs. We propose that healthcare value is created by strategic actions taken at specific points-in-time during the treatment process. Our theoretical development identifies TDC mechanisms through which IT and analytics displace later high cost interventions in favor of earlier preventative procedures. We test our hypotheses using four years of data on 45,000 cardio-metabolic patients from the U.S. state of Vermont, which implemented a Patient-Centered Medical Home (PCMH) program. Our study includes four cohorts with increasing levels of IT and analytics use: (1) non-PCMH practices, (2) PCMH practices with basic IT systems installed, (3) practices that completed data quality sprints (DQS) to increase use of IT systems, and (4) practices that use analytics through the Vermont Healthcare Information Exchange (VHIE). Our results provide insights into how TDC effects develop over time. In Year 1 after implementation, the DQS cohort demonstrates a marked increase in the use of preventative procedures such as eye exams and neuropathy screenings, the increase becomes more pronounced in Years 2 and 3, and the increase is even greater for the VHIE cohort. As the use of preventative procedures increases, emergency department utilization decreases, with a more pronounced decrease for the VHIE cohort than the DQS cohort. By Year 2, the DQS and VHIE cohorts experience a decrease in total healthcare costs, with a greater decrease for the VHIE cohort than the DQS cohort. By Year 3, the healthcare outcomes indicator of Hemoglobin A1c (HbA1c) level is statistically significantly lower, with a greater decrease for the VHIE cohort than the DQS cohort. The increased use of low-intervention healthcare treatments earlier in the process leads to a decrease in overall healthcare costs, which then leads to an improvement in healthcare indicators. Collapse

Topics: information system use health information system IS value organizational value information exchange

Methods: experimental group propensity score matching robustness check descriptive statistic qualitative interview

The Network Structure of Successful Collaboration in Wikipedia

2019 | HICSS | Citations: 0

Authors: Lerner, Jurgen; Lomi, Alessandro

Abstract: Wikipedia is one of the largest and most successful examples of decentralized pe ... Expand

Abstract: Wikipedia is one of the largest and most successful examples of decentralized peer-production systems currently in existence. Yet, the quality of Wikipedia articles varies widely with articles considered of encyclopedic quality (called featured articles) representing less than 0.1 percent of all articles. In this paper, we examine how article quality varies as a function of the network mechanisms that control the interaction among contributors. More specifically, we compare the network mechanisms underlying the production of the complete set of featured articles, with the network mechanisms of a contrasting sample of comparable non-featured articles in the English-language edition of Wikipedia. Estimates of relational event models suggest that contributors to featured articles display greater deference toward the reputation of their team members. Contributors to featured articles also display a weaker tendency to follow the behavioral norms predicted by the theory of structural balance, and hence a weaker tendency toward polarization. Collapse

Topics: Wikipedia network effect manufacturing information system team productivity

Methods: literature sample descriptive statistic case control study logistic regression

Theories: balance theory

Gauge the Effects of Targeted Advertising along the Consumer Funnel

2019 | HICSS | Citations: 0

Authors: Zhang, Ran; Zantedeschi, Daniel; Shivendu, Shivendu

Abstract: Targeted display advertising for individual consumers has become pervasive on so ... Expand

Abstract: Targeted display advertising for individual consumers has become pervasive on social media platform and other online websites (traditional platform). Yet, the effectiveness of targeted advertising across online platforms is not well understood. Moreover, such advertising effect may be different for different types of consumers, i.e. consumers in the early stage and those in the late stage, relative to the final purchase stage. This paper aims at assessing the effectiveness of targeted advertising across online platforms on consumers' final conversion (purchase). In addition, we measure the complementarity and substitutability of online platforms for targeted advertising for upper funnel (early-stage) consumers and lower funnel (late-stage) consumers. We use machine learning techniques to form case-control designs analyzed employing regularized discrete choice models to select relevant features explaining the final conversion. The empirical analysis shows that (1) targeting across platforms is positively associated with the final conversion for the lower funnel consumers, but there is no measurable synergistic effect for the upper funnel consumers; (2) the main effect of targeting on social media is positively related to the final conversion for consumers in the upper funnel but has no significant impact for lower funnel consumers. We leverage upon these findings to discuss actionable managerial prescriptions. Collapse

Topics: social media advertising management marketing management online advertising website

Methods: logistic regression experimental group computational algorithm machine learning case control study

Disintegrating Information Technology in Corporate Divestures: Implications for Regulatory Compliance Risks and Costs

2009 | International Conference on Information Systems | Citations: 9

Authors: Tanriverdi, Huseyin

Abstract: Prior research has paid very little attention, if at all, to the risks and costs ... Expand

Abstract: Prior research has paid very little attention, if at all, to the risks and costs entailed in IT disintegration processes. In this study, we begin addressing this gap by studying how IT disintegration challenges posed by corporate divestitures affect the regulatory compliance risks and costs of divesting firms in the context of the Sarbanes-Oxley Act of 2002 (SOX). We hypothesize that firms with higher corporate divestiture intensities are more likely to have material weaknesses in their IT controls, more likely to become incompliant with SOX, and more likely to incur higher auditor fees during SOX audits. We also hypothesize that superior IT capabilities could reduce the probability and magnitude of the regulatory compliance risks and costs during divestitures. We find empirical support for these hypotheses in a sample of 252 publicly traded U.S firms that were audited independently for SOX compliance between 2004 and 2008. Collapse

Topics: Sarbanes–Oxley Act information technology capability database system accounting financial performance

Methods: Compustat SDC Platinum cluster analysis theory development case control study