An official website of the United States government
Auditing social-media algorithms has become a focus of public-interest research and policymaking to ensure their fairness across demographic groups such as race, age, and gender in consequential domains such as the presentation of employment opportunities. However, such demographic attributes are often unavailable to auditors and platforms. When demographics data is unavailable, auditors commonly \emph{infer} them from other available information. In this work, we study the effects of inference error on auditing for bias in one prominent application: \emph{black-box} audit of ad delivery using \emph{paired ads}. We show that inference error, if not accounted for, causes auditing to falsely miss skew that exists. We then propose a way to mitigate the inference error when evaluating skew in ad delivery algorithms. Our method works by adjusting for expected error due to demographic inference, and it makes skew detection more sensitive when attributes must be inferred. Because inference is increasingly used for auditing, our results provide an important addition to the auditing toolbox to promote correct audits of ad delivery algorithms for bias. While the impact of attribute inference on accuracy has been studied in other domains, our work is the first to consider it for black-box evaluation of ad delivery bias, when only aggregate data is available to the auditor.
more »
« less
Understanding Practices, Challenges, and Opportunities for User-Engaged Algorithm Auditing in Industry Practice
Recent years have seen growing interest among both researchers and practitioners in user-engaged approaches to algorithm auditing, which directly engage users in detecting problematic behaviors in algorithmic systems. However, we know little about industry practitioners’ current practices and challenges around user-engaged auditing, nor what opportunities exist for them to better leverage such approaches in practice. To investigate, we conducted a series of interviews and iterative co-design activities with practitioners who employ user-engaged auditing approaches in their work. Our findings reveal several challenges practitioners face in appropriately recruiting and incentivizing user auditors, scaffolding user audits, and deriving actionable insights from user-engaged audit reports. Furthermore, practitioners shared organizational obstacles to user-engaged auditing, surfacing a complex relationship between practitioners and user auditors. Based on these findings, we discuss opportunities for future HCI research to help realize the potential (and mitigate risks) of user-engaged auditing in industry practice.
more »
« less
- Award ID(s):
- 2040942
- PAR ID:
- 10573770
- Publisher / Repository:
- ACM
- Date Published:
- ISBN:
- 9781450394215
- Page Range / eLocation ID:
- 1 to 18
- Format(s):
- Medium: X
- Location:
- Hamburg Germany
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Purpose This conversation presents the reflections from five prominent disaster scholars and practitioners on the opportunities and challenges associated with research following disasters and explores the importance of ethics in disaster research. Design/methodology/approach This paper is based on the conversations that took place on Disasters: Deconstructed Podcast livestream on the 11th of June 2021. Findings The prominent themes in this conversation include ethical approaches to research, how we–as disaster researchers and practitioners–collaborate, engage, and cooperate, and whose voices are centred in a post-disaster research context. Originality/value The conversation contributes to ongoing discussions around the conduct and practice of disaster research.more » « less
-
An emerging body of research indicates that ineffective cross-functional collaboration – the interdisciplinary work done by industry practitioners across roles – represents a major barrier to addressing issues of fairness in AI design and development. In this research, we sought to better understand practitioners’ current practices and tactics to enact cross-functional collaboration for AI fairness, in order to identify opportunities to support more effective collaboration. We conducted a series of interviews and design workshops with 23 industry practitioners spanning various roles from 17 companies. We found that practitioners engaged in bridging work to overcome frictions in understanding, contextualization, and evaluation around AI fairness across roles. In addition, in organizational contexts with a lack of resources and incentives for fairness work, practitioners often piggybacked on existing requirements (e.g., for privacy assessments) and AI development norms (e.g., the use of quantitative evaluation metrics), although they worry that these tactics may be fundamentally compromised. Finally, we draw attention to the invisible labor that practitioners take on as part of this bridging and piggybacking work to enact interdisciplinary collaboration for fairness. We close by discussing opportunities for both FAccT researchers and AI practitioners to better support cross-functional collaboration for fairness in the design and development of AI systems.more » « less
-
Obonyo, Esther (Ed.)The construction industry is rapidly changing due to the greater adoption of innovations and technology. This has necessitated changes in the competencies that the industry demands from new graduates. For academia to meet the changing needs of the industry, the inputs of practitioners are needed to complement academic pedagogical efforts. This study leverages the potential of Web 2.0 to develop a web platform called ConPEC to facilitate instructor-practitioner collaborations for enhancing student learning. ConPEC is aimed at providing instructors with equitable access to practitioners, increasing the participation of practitioners in instructors' pedagogical efforts, and enabling greater interaction of students with their communities of practice (CoP). These could facilitate achieving a proper blend of theory and practice in construction engineering education as well as ensure that students possess the competencies that the industry demands. This study demonstrates the efficacy of design principles in designing information systems. This study also demonstrates the usage of the Technology Acceptance Model (TAM) to explain and understand practitioners' acceptance of ConPEC. The findings reveal that practitioners perceived ConPEC to be useful, easy to use, and user-friendly. Practitioners’ behavioral intention-to-use ConPEC is significantly influenced by attitude toward usage, perceived ease of use, and trust. Trust also significantly influenced perceived ease of use. However, perceived usefulness has no direct significant influence on practitioners’ behavioral intention-to-use ConPEC. The study uncovers practitioners' acceptance behavior toward ConPEC which could be leveraged for further system development. The study also provides a framework that can be leveraged in diverse domains to develop similar initiatives aimed at addressing skill gaps in fresh graduates.more » « less
-
Engineering education researchers have identified a lack of alignment between the complexities of professional engineering contexts and the ways that we train and evaluate the ethical abilities and dispositions of engineers preparing for professional practice. The challenges that engineers face as practitioners are multifaceted, wicked problems situated in unique and varied disciplinary and industry contexts. Understanding the variations in ways of experiencing ethics by practicing engineers in these complex professional contexts will support a better alignment between engineering ethics instruction and what students might experience in professional practice. While there is a need for richer and more contextually-specific ethics training for many areas, our initial focus is the healthcare products industry. Thus, our NSF-funded CCE STEM project will enable us to analyze the alignment of relationships among frameworks for ethics education in engineering and the reality of engineering practice within the health products industry. As a first phase, the project has focused on understanding the different ways in which practicing engineers experience ethical issues in the health products industry using phenomenography, an empirical research methodology for investigating qualitatively different ways people experience a phenomenon. In the second phase, we have analyzed critical incidents that potentially cause the variation in experiencing ethics in practice. The findings of these studies are anticipated to serve as a guidepost for aligning educational strategies and developing effective training for future ethical practitioners. In our paper, we present an overview of the study (background and methods), progress to date, and how we expect the results to inform engineering ethics education and industry ethics training.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1145/3544548.3581026
