Babak Safaei

Research

Working Papers

  • Co-authors: Babak Safaei, Milena Head

  • Status: Targeting Management Information Systems Quarterly (MISQ)

  • Selected Presentations: Digital Futures Symposium (2025), Americas Conference on Information Systems (AMCIS) Doctoral Consortium (2025), Americas Conference on Information Systems (AMCIS) (2025), Pre ICIS SIGHCI Workshop (2025), Americas Conference on Information Systems (AMCIS) (2024).

This research investigates the situational effectiveness of fact-checking interventions aimed at mitigation misinformation on social media platforms. It specifically examines how a fact-checker’s source, such as an AI system, a human expert, or a crowd consensus, influences user credibility judgments. The study also explores how these perceptions differ across personal beliefs and user age groups, critical factors given the varying vulnerability to online deception. Using a large scale online experiment, this paper develops and tests a theoretical model of how users perceive objectivity and agency in corrective labels considering the source-related heuristics. The goal is to provide evidence based guidance for digital platforms on designing effective, age responsive content moderation interventions that mitigate harm and support adaptive user behaviors.

  • Co-authors: Babak Safaei, Milena Head

  • Status: Targeting Information Systems Research (ISR)

  • Selected Conferences: Americas Conference on Information Systems (AMCIS) (2024), Pre ICIS SIGHCI Workshop (2023), Pre ICIS SIGSEC Workshop (2023).

This project examines the factors that contribute to phishing susceptibility, with a particular focus on differences across the user lifespan. Anchored in dual process and prospect theory, the study develops a model incorporating individual traits, cognitive states, and message framing (e.g., gain versus loss). This research employs a NeuroIS methodology, combining behavioral survey data with eye tracking and physiological measurements (EDA) to capture process level evidence of cognitive and emotional engagement. The findings aim to inform the design of safer interfaces and personalized security awareness programs.

  • Co-authors: Babak Safaei, Mahdi Mirhosseini, Shamed Addas, Milena Head

  • Status: Experimental Design Phase; Targeting Information Systems Research (ISR)

This study complements research on misinformation by examining the mechanisms of feedback delivery. Integrating feedback intervention theory with signal detection theory, this project decomposes user performance into detection sensitivity and response bias. Through a factorial online experiment, the research tests how different types of feedback (e.g., corrective, empathetic) and their presentation influence a user’s ability to discern true content from false content. The objective is to provide actionable guidance on how to design feedback interventions that effectively improve user accuracy.

Working in Progress

  • Co-authors: Bahar Majd, Babak Safaei, Motahareh (Bahar) Pourbehzadi, Giti Javidi

  • Status: Data Analysis Phase; Targeting Journal of Business Ethics

This paper investigates how leading AI firms construct and communicate ethical responsibility through their public disclosures on AI ethics and cloud security. It asks whether such narratives represent genuine ethical governance or forms of “ethics-washing” aimed at maintaining legitimacy. Drawing on Goffman’s dramaturgy, legitimacy theory, and institutional decoupling, the study interprets corporate ethics communication as a performance of legitimacy, revealing the tension between symbolic and substantive action. It introduces a novel framework of literary genres to analyze how companies stylistically present ethical identity across cultural and regulatory contexts. The study contributes to business ethics and information systems scholarship by reframing ethics-washing as both an organizational and cultural phenomenon, integrating socio-legal, performative, and aesthetic dimensions of corporate AI governance.

  • Co-authors: Babak Safaei, Srinivas Sekar

  • Status: Writing in Progress; Targeting NeurIPS 2026

This methodological paper, developed from my industry experience at Capital One, introduces a novel framework for feature selection in high-stakes financial technology machine learning risk models. The framework combines game-theoretic principles (SHAP, SAGE) with recursive feature elimination to improve model parsimony and interpretive stability for gradient-boosted risk models. This approach is designed to strengthen governance and auditability, enabling human oversight and calibration of trust in Human-AI collaboration contexts of automated financial decision making.

  • This stream includes two scoping reviews conducted as part of the EMPOWrD (Enhancing Mobility and Participation for Older Adult Wellness through Digital Inclusion) research program. I played an active role in this program’s foundation by contributing to the grant development and research planning. Within this program, I lead a meta-analysis quantifying the effectiveness of content-level accuracy signals on user trust in misinformation contexts, with moderators for age strata and AI provenance disclosures. I also co-lead two scoping reviews: one on technology-mediated social connection for older adults living with dementia, and another on the integration of HCI and software engineering practices for aging-in-place.

  • Project 1: AI Disclosures and Fact-Check Provenance in Misinformation Contexts: A Meta-Analysis of User Responses

    • Co-authors: Babak Safaei, Milena Head

    • Status: Data Collection Phase; Targeting Information & Management

  • Project 2: Technology-Mediated Social Connection for Older Adults Living with Dementia: A Scoping Review.

    • Co-authors: Babak Safaei, Irina Ghilic, Constance Dupuis, Nicole Dalmer, Sheila Boamah

    • Status: Data Collection Phase

  • Project 3: Synthesizing HCI, Requirements Engineering, and Design Science for Aging-In-Place Software Development: A Scoping Review.

    • Co-authors: B. Safaei, P. J. White, Irina Ghilic, Shyam Ravichandran, Denise Geiskkovitch, Sebastien Mosser

    • Status: Data Collection Phase

This research employs a historical-literary and comparative methodology to analyze leadership and organizational governance in Attar Neyshaburi’s 12th-century Persian epic, The Conference of the Birds. The study interprets the flock’s arduous journey as an organizational allegory, examining how their leader, the Hoopoe, uses sophisticated management strategies, such as personalized motivation, individualized leader-member exchanges, and realistic risk previews, to maintain commitment and organizational synergy. We demonstrate that these implicit theories of leadership align closely with contemporary management frameworks. My experience with historiography and comparative textual analysis sharpens my conceptualization of authority, legitimacy, and governance and equips me with a unique lens to examine the evolution of modern socio-technical phenomena in Information Systems.

  • Co-authors: Babak Safaei, Bahar Majd, Vishwanath Baba

  • Status: Major Revision (First Round) at Journal of Management History (JMH). 

Published Academic Papers

Published (Information Systems and Management)

  • Safaei, B., & Head, M. (2025). The messenger matters: How fact-checker identity and user age shape credibility judgments online. Proceedings of the 24th Annual Pre-ICIS Workshop on HCI Research in MIS (SIGHCI). [Forthcoming]
  • Safaei, B., & Head, M. (2025). Fact-checking misinformation: Human-AI collaboration dynamics across age groups. Americas Conference on Information Systems (AMCIS), 196.
  • Safaei, B., & Head, M. (2024). Navigating the digital frontier: AI and coping mechanisms for enhancing misinformation resilience in older adults. Americas Conference on Information Systems (AMCIS), 160.
  • Safaei, B., Yuan, Y., & Safaei, N. (2024). Empowering organizations through big data: A framework for digital resilience. Americas Conference on Information Systems (AMCIS), 161.
  • Safaei, B., & Head, M. (2024). Investigating age-related factors in phishing susceptibility: A focus on decision-making processes in the HCI context. Proceedings of the 22nd Annual Pre-ICIS Workshop on HCI Research in MIS (SIGHCI), 5.
  • Safaei, B., Majd, B., & Baba, V. (2023). The Conference of the Birds: Lessons in leadership and management from 12th-century Persia. Proceedings of the Administrative Sciences Association of Canada (ASAC).
  • Safaei, N., Safaei, B., Seyedekrami, S., Talafidaryani, M., Masoud, A., Wang, S., Li, Q., & Moqri, M. (2022). E-CatBoost: An efficient machine learning framework for predicting ICU mortality using the eICU collaborative research database. PLOS One, 17(5), e0262895.

Published (Engineering)

  • Mahmud, M. S., Gates, T. J., Savolainen, P. T., & Safaei, B. (2023). Driver response to a dynamic speed feedback sign at a freeway exit ramp, considering the sign design and installation characteristics. Transportation Research Record, 2677(3), 289-301.
  • Safaei, N., Smadi, O., Masoud, A., & Safaei, B. (2022). An automatic image processing algorithm based on crack pixel density for pavement crack detection and classification. International Journal of Pavement Research and Technology, 15(1), 159-172.
  • Safaei, B., Safaei, N., Masoud, A., & Seyedekrami, S. (2021). Weighing criteria and prioritizing strategies to reduce motorcycle-related injuries using a combination of fuzzy TOPSIS and AHP methods. Advances in Transportation Studies, 54.
  • Safaei, N., Smadi, O., Safaei, B., & Masoud, A. (2021). Efficient road crack detection based on an adaptive pixel-level segmentation algorithm. Transportation Research Record, 2675(9), 370-381.
  • Mahmud, M. S., Gupta, N., Safaei, B., Jashami, H., Gates, T. J., Savolainen, P. T., & Kassens-Noor, E. (2021). Evaluating the impacts of speed limit increases on rural two-lane highways using quantile regression. Transportation Research Record, 2675(11), 740-753.
  • Safaei, N., Zhou, C., Safaei, B., & Masoud, A. (2021). Gasoline prices and their relationship to the number of fatal crashes on US roads. Transportation Engineering, 4, 100053.

Skills and Software

  • Quantitative & Econometric Analysis (SPSS, AMOS, Stata, R, JASP)
  • Python Stats (statsmodels)
  • Machine Learning & Explainable AI (Python: scikit-learn, Gradient Boosting: XGBoost, LightGBM, CatBoost, GLMs; feature selection, parsimony analytics; XAI: SHAP, SAGE, permutation)
  • Data Engineering & Analytics (Python: Pandas, Numpy, SciPy; Platforms: PySpark, Databricks, SQL; MLOps, Pipelines: Kubernetes, Kubeflow; Visualization: Tableau, Power BI, seaborn, Matplotlib)
  • NeuroIS & Physiological Computing (Python: MNE, NeuroKit2; Software: BIOPAC AcqKnowledge, Noldus Observer)
  • Deep Learning, Natural Language Processing, Computer Vision (Python: PyTorch, TensorFlow, HuggingFace Transformers, OpenCV, Torch-Text)
  • Qualitative Data Analysis (NVivo, Dedoose, Elicit)