Introduction to the Minitrack on Digital Methods

2023 | HICSS | Citations: 0

Authors: Walker, S.; Halavais, A.; Weber, M.

Abstract: ... Expand

Abstract: Collapse

Topics: social media affordance application programming interface Twitter

Methods: digital trace data

Patient health locus of control: the design of information systems for patient-provider interactions

2023 | European Journal of Information Systems | Citations: 0

Authors: Wallace, James; Mullarkey, Matthew T; Hevner, Alan

Abstract: Patient locus of control (LOC) is a strong determinant of health outcomes, yet h ... Expand

Abstract: Patient locus of control (LOC) is a strong determinant of health outcomes, yet healthcare systems rarely use patient LOC to devise treatment plans. Our clinical research applies action design research methods to design and evaluate healthcare information systems to improve paths and outcomes of patient care. Socio-technical synergies between the interior (technical) and exterior (socio) designs support effective human-computer interfaces and interactions. Rigorous focus group evaluations of the systems for capturing LOC information and using that information in treatment plans are performed and system refinements are implemented. The practitioner organisation has future plans for full implementation of the designed systems. Collapse

Topics: health information system systems design self efficacy

Methods: focus group action design research survey design science qualitative interview

Gamification and policy compliance: Results from an online vignette experiment in the context of social distancing for public health security

2023 | HICSS | Citations: 0

Authors: Wallius, E.; Tome Klock, A.; Hamari, J.

Abstract: The commonly applied strategies for promoting compliance with public health and ... Expand

Abstract: The commonly applied strategies for promoting compliance with public health and safety policies can be inefficient and coercive, posing a need to examine novel motivational strategies to aid in this endeavor. Gamification, which aims to foster engagement and intrinsic motivation towards mundane activities and behaviors, is one of the vanguard design approaches among behavioral change support systems. Despite the increasing interest in gamification, the corpus lacks studies on its effects on policy compliance. Therefore, this study examines the relationships between gamification design types, gameful experience, and policy compliance in the social distancing context (during COVID-19) using a vignette-based online experiment (n=937). Based on the results, gameful experience mediates the positive relationships between achievement and progression-based, competitive, and immersive gamification and policy compliance, while social gamification is not associated with gameful experience. The results provide evidence of gamification's potential as a non-coercive method of helping people follow policies. Collapse

Topics: gamification pandemic usability user interface IT work

Methods: experiment survey case study PLS tool psychometrics

Taking Advantage of Algorithmic Preference to Reduce Product Returns in E-Commerce

2023 | International Conference on Information Systems | Citations: 0

Authors: Walsh, Gianfranco; Schaarschmidt, Mario

Abstract: Reimbursement of repair costs is a way to motivate customers to keep defective ... Expand

Abstract: Reimbursement of repair costs is a way to motivate customers to keep defective products instead of returning them. However, there is no research-based guidance on how retailers should frame repair costs reimbursement offers in terms of who decides on the size of the reimbursement and makes the offer-an employee or a machine. To guide further IS research and suggest ways that help e-commerce businesses to improve repair costs reimbursement effectiveness to decrease product return rates, the present research draws on literature on offer sources and on insights from a qualitative and an experimental study. We find that artificial intelligence-based (vs. human-based) repair costs reimbursement offers promote fairness perceptions, which, in turn, affect important customer outcomes-the likelihood to accept the offer and digital negative word of mouth. The results can guide e-commerce businesses' returns-prevention efforts and IS research. Collapse

Topics: artificial intelligence electronic commerce word of mouth decision making crowdsourcing

Methods: experiment survey design qualitative interview business process modeling mediation analysis

Theories: equity theory distributive justice theory

BIBGT: combining bibliometrics and grounded theory to conduct a literature review

2023 | European Journal of Information Systems | Citations: 0

Authors: Walsh, Isabelle; Rowe, Frantz

Abstract: In the current context of scientific information overload, we propose a method c ... Expand

Abstract: In the current context of scientific information overload, we propose a method combining bibliometrics and grounded theory to conduct literature reviews that have a descriptive, understanding or explanatory purpose. This overall inductive combined method, which we name BIBGT (BIB = Bibliometrics; GT = Grounded Theory) provides a powerful instrument for researchers. This instrument has been made more readily accessible by recent technological developments, and scientific advances in the field of bibliometrics. Notably, BIBGT helps grounded theorists identify colleges of thought of the field being reviewed, and representative texts to be reviewed in depth, in each college; it improves the trustworthiness and efficiency of reviews. BIBGT helps bibliometricians achieve in-depth analysis and interpretation of their findings. For all researchers, BIBGT can nurture the emergence of novel insights towards in-depth description, understanding or explanation of a core concept, theme, or research domain.We detail the four steps of this flexible method, providing essential elements and some guidelines as well as an illustration, to help other researchers who wish to apply it. Collapse

Topics: IT alignment database system information overload

Methods: bibliometric analysis literature study qualitative coding co-citation analysis grounded theory

Boosting Benefits, Offsetting Obstacles – the Impact of Explanations on AI Users’ Task Performance

2023 | International Conference on Business Informatics | Citations: 0

Authors: Walter, Marie Christine; Broder, Hanna Rebecca; Förster, Maximilian

Abstract: Artificial Intelligence (AI) bears the potential to inform human decision-makin ... Expand

Abstract: Artificial Intelligence (AI) bears the potential to inform human decision-making in a large variety of domains. However, its "black box" character poses an obstacle to human agency in interaction with AI-based decision support. A possible solution comes from the research field of Explainable AI (XAI), which generates explanations that reveal AI's functioning to users. Our research on XAI focuses on understanding the immediate and prolonged effect of XAIbased decision support on task performance. To this end, we conducted a randomized between-subjects online experiment with 289 participants performing the task of image classification. We find that explanations along AI decisions boost the positive effect of AI-based decision support on task performance during interaction. Furthermore, explanations can counterbalance the potential negative effect on prolonged task performance, which manifests after AI-based decision support is being withdrawn. Our findings contribute to understanding the impact of XAI on the outcome of human-AI interaction. Collapse

Topics: artificial intelligence task productivity decision support explainable artificial intelligence decision support system

Methods: experimental group experiment Mann Whitney U test nonparametric test statistical hypothesis test

Doing Well by Doing Fair? Investigating the Effect of Users’ Fairness Perceptions on Adoption Intentions of Services

2023 | International Conference on Information Systems | Citations: 0

Authors: Waltermann, Hubertus; Ziegler, Anika; Engert, Simon; Hess, Thomas

Abstract: Within the IS discipline, the concept of fairness as a determinant for social g ... Expand

Abstract: Within the IS discipline, the concept of fairness as a determinant for social good recently gained attention. However, a comprehensive understanding of the influence of different fairness dimensions on user perceptions and their adoption intention of new technologies is missing. Based on fairness research, the FAIRSERV model, and the technology acceptance model, we derived a research model to study how users' fairness perceptions along different dimensions influence their adoption intentions in the context of online services. We tested our model in an online experiment with 407 participants. Our results show that perceived distributive, procedural, and interactional fairness positively influence users' perceived overall fairness regarding an online service, which, in return, positively influences their adoption intentions. We contribute to IS research by shedding light on the societal impact of fairness in the context of adopting new digital services. Practitioners can utilize our findings to improve their service offerings. Collapse

Topics: privacy societal evaluation criteria digital service user acceptance information system use

Methods: experiment factor analysis partial least squares path modeling survey chi squared test

Theories: technology acceptance model equity theory

Division of Labor between Humans and Algorithms in Healthcare: The Case of Surgery Duration Predictions

2023 | International Conference on Information Systems | Citations: 0

Authors: Walzner, Dominik; Poreschack, Laura Maria; Fuegener, Andreas; Schiffels, Sebastian; Denz, Christof

Abstract: For many healthcare applications a collaboration of humans and algorithms has b ... Expand

Abstract: For many healthcare applications a collaboration of humans and algorithms has been shown to be superior to pure automation in terms of performance. However, the healthcare sector is characterized by shortages in personnel, which can lead to an excessive workload for the employees and thus makes automation highly beneficial to reduce human workload. In our paper, we consider a combination of different work modes and evaluate whether humans have to be involved in every instance of a task or whether they can be replaced by an AI for some instances. We analyze the potential of segmenting tasks based on who is involved in their completion: Either an AI or a human complete the task individually, or they complete the task together. Considering the case of surgery duration predictions and using a dataset from a university hospital, we observe that human effort could be decreased while maintaining a high prediction performance. Collapse

Topics: artificial intelligence human computer interaction human AI collaboration task productivity

Methods: computational algorithm decision tree classification random forest classification machine learning case study

Improving triaging from primary care into secondary care using heterogeneous data-driven hybrid machine learning

2023 | Decision Support Systems | Citations: 0

Authors: Wang, Bing; Li, Weizi; Bradlow, Anthony; Bazuaye, Eghosa; Chan, Antoni T.Y.

Abstract: Effective and rapid triaging from primary care into secondary care plays a pivot ... Expand

Abstract: Effective and rapid triaging from primary care into secondary care plays a pivotal role in providing patients with timely treatment and managing increasing demands for healthcare resources. Existing triaging methods from primary care to secondary care are labor-intensive processes that involve manually reviewing referral data from multiple sources and can cause long referral to treatment time. There has been no research using machine learning methods that automatically analyzes heterogeneous data including referral letters to recognize regu­ larities to support the primary to secondary care triage. In this paper, we propose a heterogeneous data-driven hybrid machine learning model including Natural Language Processing (NLP) to improve hospital triage effi­ ciency at the point of triage. The proposed model achieved a precision of 0.83, recall of 0.82, F1-Score of 0.83, accuracy of 0.82, AUC of 0.90 in identifying patients with non-inflammatory conditions (NIC) and inflammatory arthritis (IA) at the point of triage with explainable risk stratifications. Our model is piloted in a real-world trial in a large secondary care hospital in the UK to compare referral accuracy and time saved between our model and clinicians, and evaluate its acceptability by users. Our model achieved precision and recall of 0.83 and 0.81, compared with the precision and recall of 0.80 and 0.78 by clinicians. The research also shows that our model enabled decision support can save clinicians 8 h per week in assessing the referral assessment. This paper is the first study to streamline hospital triage from primary care to secondary care using machine learning. Collapse

Topics: local interpretable model agnostic explanation decision support decision support system data governance data protection

Methods: machine learning BERT data imputation natural language processing ensemble learning

Social Movements, Collective Identity, and Workplace Allies: The Labeling of Gender Equity Policy Changes

2023 | Organization Science | Citations: 1

Authors: Wang, Cynthia S.; Whitson, Jennifer A.; King, Brayden G; Ramirez, Rachel L.

Abstract: Social movements seek allies as they campaign for social, political, and organiz ... Expand

Abstract: Social movements seek allies as they campaign for social, political, and organizational changes. How do activists gain allies in the targeted institutions they hope to change? Despite recognition of the importance of ally support in theories about institutional change and social movements, these theories are largely silent on the microdynamics of ally mobilization. We examine how the labeling of organizational policies that benefit women influences potential workplace allies’ support for these policies. We theorize that one barrier to mobilizing workplace allies is a misalignment of the labels that activists use to promote new policies and employees’ affiliation with collective identities. We conducted five experiments to test our hypotheses and 26 qualitative interviews to provide illustration of our core concepts. We demonstrate that employees high in feminist identification are more likely to support feminist-labeled (feminist and #MeToo) than unlabeled policies, whereas those low in feminist identification are less likely to support feminist-labeled than unlabeled policies (Studies 1–3). However, we find that participants for whom organizational identification was high (whether measured or manipulated) and feminist identification was low supported organizationally labeled policies more than feminist-labeled polices (Studies 4 and 5). This illustrates that policies whose aims may not align with one collective identity can still garner support by activating another relevant collective identity. Within our studies, we provide evidence that these effects are mediated via feelings of pride in the organization (and not fear or anger), suggesting that positive emotions are a central mechanism in mobilizing workplace allies.History: This paper has been accepted for the Organization Science Special Issue on Experiments in Organizational Theory.Supplemental Material: The online appendix is available at https://doi.org/10.1287/orsc.2021.1492. Collapse

Topics: marketing management research and development organizational context accounting legal context

Methods: qualitative interview theory development experiment linear regression analysis survey

Theories: organizational theory reinforcement sensitivity theory social identity theory