Eddie Murphy and the Dangers of Counterfactual Causal Thinking About Detecting Racial Discrimination

By Issa Kohler-Hausmann. 2019

123 citations recorded by the source corpus.

Subfields: Political Science and International Relations, Sociology and Political Science, Causal Inference & Discovery, Social Sciences & Humanities

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Political Science and International Relations

In Sociology and Political Science

In Causal Inference & Discovery

In Social Sciences & Humanities

Citations in the corpus

Ethical and social risks of harm from Language Models (2021, background): “Determining what constitutes a truly fair and equitable training dataset may also require further research in Ethics and Law (Kohler-Hausmann, 2019).”
Counterfactual Fairness in Text Classification through Robustness (2018, background): “ly intervene on identity of an individual; given that most social scientists agree that race is “socially constructed,” it may be unreasonable to attempt to modify race and all its associated factors [10]. These limitations, among others, are reflected in debate surrounding the use of counterfactual”
Fairness and Abstraction in Sociotechnical Systems (2019, neutral): “[45, p. 7].”
Towards a critical race methodology in algorithmic fairness (2019, neutral): “mselves contribute to the essentialization of race [10, 42, 43, 51]. There has also been signiﬁcant debate about the appropriate use of racial variables in studies aimed at identifying causal effects [46, 54, 55, 58, 67]. Objections to the use of race as a causal variable are often couched in terms ”
Towards Intersectionality in Machine Learning: Including More Identities, Handling Underrepresentation, and Performing Evaluation (2022, background): “Their permissibility of use, however, for addressing class imbalance has been contested because of the infeasibility of manipulating demographic categories [62, 74, 110] and inaccurate conceptualizations of causality of demographic categories [64, 83].”
Image Counterfactual Sensitivity Analysis for Detecting Unintended Bias (2019, background): “Recently, [23, 29] have argued this framework is based on a misunderstanding of how race operates in our society and the ﬂawed assumption that social groups like race and gender are entities that can have counterfactual causality.”
This Thing Called Fairness (2019, background): “In particular, if one subscribes to a constructivist theory of the attributes that define membership in privileged and underprivileged groups, it seems inherently problematic to consider these factors as “attributes” [80] Similarly the concept of intersectionality, which identifies the ways that rac”
The fallacy of inscrutability (2018, background): “Because systems are designed to achieve particular ends, we cannot disentangle the environmental contributions to a particular behaviour from the designed mechanism [31]—the mechanism must, after all, compensate for its environment if it correctly embodies its goals.”
Algorithmic Fairness: Choices, Assumptions, and Definitions (2021, methodology): “We define random variables V ( a ), D ( a ) as values randomly drawn from the population (Kohler-Hausmann 2018).”
Extending the Machine Learning Abstraction Boundary: A Complex Systems Approach to Incorporate Societal Context (2020, background): “7SCMs are optimized for bridging the gap between qualitative and quantitative description and to explicitly deal with the fact that data alone cannot be utilized to understand the underlying causal structures that generated it [76] 8Critiques of counterfactual approaches caution that treating concep”