-
Beyond Recommendations: From Backward to Forward AI Support of Pilots' Decision-Making Process
Authors:
Zelun Tony Zhang,
Sebastian S. Feger,
Lucas Dullenkopf,
Rulu Liao,
Lukas Süsslin,
Yuanting Liu,
Andreas Butz
Abstract:
AI is anticipated to enhance human decision-making in high-stakes domains like aviation, but adoption is often hindered by challenges such as inappropriate reliance and poor alignment with users' decision-making. Recent research suggests that a core underlying issue is the recommendation-centric design of many AI systems, i.e., they give end-to-end recommendations and ignore the rest of the decisi…
▽ More
AI is anticipated to enhance human decision-making in high-stakes domains like aviation, but adoption is often hindered by challenges such as inappropriate reliance and poor alignment with users' decision-making. Recent research suggests that a core underlying issue is the recommendation-centric design of many AI systems, i.e., they give end-to-end recommendations and ignore the rest of the decision-making process. Alternative support paradigms are rare, and it remains unclear how the few that do exist compare to recommendation-centric support. In this work, we aimed to empirically compare recommendation-centric support to an alternative paradigm, continuous support, in the context of diversions in aviation. We conducted a mixed-methods study with 32 professional pilots in a realistic setting. To ensure the quality of our study scenarios, we conducted a focus group with four additional pilots prior to the study. We found that continuous support can support pilots' decision-making in a forward direction, allowing them to think more beyond the limits of the system and make faster decisions when combined with recommendations, though the forward support can be disrupted. Participants' statements further suggest a shift in design goal away from providing recommendations, to supporting quick information gathering. Our results show ways to design more helpful and effective AI decision support that goes beyond end-to-end recommendations.
△ Less
Submitted 13 June, 2024;
originally announced June 2024.
-
Towards Sustainable Research Data Management in Human-Computer Interaction
Authors:
David Goedicke,
Mark Colley,
Sebastian S. Feger,
Michael Goedicke,
Bastian Pfleging,
Wendy Ju
Abstract:
We discuss important aspects of HCI research regarding Research Data Management (RDM) to achieve better publication processes and higher reuse of HCI research results. Various context elements of RDM for HCI are discussed, including examples of existing and emerging infrastructures for RDM. We briefly discuss existing approaches and come up with additional aspects which need to be addressed. This…
▽ More
We discuss important aspects of HCI research regarding Research Data Management (RDM) to achieve better publication processes and higher reuse of HCI research results. Various context elements of RDM for HCI are discussed, including examples of existing and emerging infrastructures for RDM. We briefly discuss existing approaches and come up with additional aspects which need to be addressed. This is to apply the so-called FAIR principle fully, which -- besides being findable and accessible -- also includes interoperability and reusability. We also discuss briefly the kind of research data types that play a role here and propose to build on existing work and involve the HCI scientific community to improve current practices.
△ Less
Submitted 19 July, 2023;
originally announced July 2023.
-
Interactive Tools for Reproducible Science -- Understanding, Supporting, and Motivating Reproducible Science Practices
Authors:
Sebastian Stefan Feger
Abstract:
Reproducibility should be a cornerstone of science as it enables validation and reuse. In recent years, the scientific community and the general public became increasingly aware of the reproducibility crisis, i.e. the wide-spread inability of researchers to reproduce published work, including their own. Scientific research is increasingly focused on the creation, observation, processing, and analy…
▽ More
Reproducibility should be a cornerstone of science as it enables validation and reuse. In recent years, the scientific community and the general public became increasingly aware of the reproducibility crisis, i.e. the wide-spread inability of researchers to reproduce published work, including their own. Scientific research is increasingly focused on the creation, observation, processing, and analysis of large data volumes. On the one hand, this transition towards computational and data-intensive science poses new challenges for research reproducibility and reuse. On the other hand, increased availability and advances in computation and web technologies offer new opportunities to address the reproducibility crisis. This thesis reports on user-centered design research conducted at CERN, a key laboratory in data-intensive particle physics.
In this thesis, we build a wider understanding of researchers' interactions with tools that support research documentation, preservation, and sharing. From a Human-Computer Interaction (HCI) perspective the following aspects are fundamental: (1) Characterize and map requirements and practices around research preservation and reuse. (2) Understand the wider role and impact of research data management (RDM) tools in scientific workflows. (3) Design tools and interactions that promote, motivate, and acknowledge reproducible research practices. Research reported in this thesis represents the first systematic application of HCI methods in the study and design of interactive tools for reproducible science. We advocate the unique role of HCI in supporting, motivating, and transforming reproducible research practices through the design of tools that enable effective RDM. This thesis paves new ways for interaction with RDM tools that support and motivate reproducible science.
△ Less
Submitted 10 November, 2020;
originally announced December 2020.
-
Designing for Reproducibility: A Qualitative Study of Challenges and Opportunities in High Energy Physics
Authors:
Sebastian S. Feger,
Sünje Dallmeier-Tiessen,
Albrecht Schmidt,
Paweł W. Woźniak
Abstract:
Reproducibility should be a cornerstone of scientific research and is a growing concern among the scientific community and the public. Understanding how to design services and tools that support documentation, preservation and sharing is required to maximize the positive impact of scientific research. We conducted a study of user attitudes towards systems that support data preservation in High Ene…
▽ More
Reproducibility should be a cornerstone of scientific research and is a growing concern among the scientific community and the public. Understanding how to design services and tools that support documentation, preservation and sharing is required to maximize the positive impact of scientific research. We conducted a study of user attitudes towards systems that support data preservation in High Energy Physics, one of science's most data-intensive branches. We report on our interview study with 12 experimental physicists, studying requirements and opportunities in designing for research preservation and reproducibility. Our findings suggest that we need to design for motivation and benefits in order to stimulate contributions and to address the observed scalability challenge. Therefore, researchers' attitudes towards communication, uncertainty, collaboration and automation need to be reflected in design. Based on our findings, we present a systematic view of user needs and constraints that define the design space of systems supporting reproducible practices.
△ Less
Submitted 14 March, 2019;
originally announced March 2019.
-
Gamification in Science: A Study of Requirements in the Context of Reproducible Research
Authors:
Sebastian S. Feger,
Sünje Dallmeier-Tiessen,
Paweł W. Woźniak,
Albrecht Schmidt
Abstract:
The need for data preservation and reproducible research is widely recognized in the scientific community. Yet, researchers often struggle to find the motivation to contribute to data repositories and to use tools that foster reproducibility. In this paper, we explore possible uses of gamification to support reproducible practices in High Energy Physics. To understand how gamification can be effec…
▽ More
The need for data preservation and reproducible research is widely recognized in the scientific community. Yet, researchers often struggle to find the motivation to contribute to data repositories and to use tools that foster reproducibility. In this paper, we explore possible uses of gamification to support reproducible practices in High Energy Physics. To understand how gamification can be effective in research tools, we participated in a workshop and performed interviews with data analysts. We then designed two interactive prototypes of a research preservation service that use contrasting gamification strategies. The evaluation of the prototypes showed that gamification needs to address core scientific challenges, in particular the fair reflection of quality and individual contribution. Through thematic analysis, we identified four themes which describe perceptions and requirements of gamification in research: Contribution, Metrics, Applications and Scientific practice. Based on these, we discuss design implications for gamification in science.
△ Less
Submitted 6 March, 2019;
originally announced March 2019.
