pith. sign in

arxiv: 2405.13693 · v4 · submitted 2024-05-22 · 💻 cs.LG

Mutatis Mutandis: Revisiting the Comparator in Discrimination Testing

Jose M. Alvarez , Salvatore Ruggieri This is my paper

Pith reviewed 2026-05-24 00:49 UTC · model grok-4.3

classification 💻 cs.LG
keywords discrimination testingcomparatorceteris paribusmutatis mutandiscausal modelingmachine learningfairness
0
0 comments X

The pith

Discrimination testing can use a mutatis mutandis comparator that models the counterfactual state of the complainant without the protected attribute's effects on other attributes.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that deriving a comparator for discrimination testing inherently requires causal modeling of how a protected attribute such as race or gender influences other traits. It distinguishes the standard ceteris paribus comparator, which seeks a pair differing only in the protected attribute, from the mutatis mutandis comparator, which instead constructs the 'what would have been' profile after removing those effects. This matters because the mutatis mutandis version permits dissimilarity across multiple attributes and thereby shifts the evidentiary basis for claims of individual discrimination. A sympathetic reader would care because most existing testing tools rest on the ceteris paribus ideal, and adopting the alternative opens concrete roles for machine learning to estimate the required adjustments from data.

Core claim

The paper claims that the complainant-comparator pair is central to discrimination testing and that the comparator's construction is a causal modeling task. It defines the mutatis mutandis comparator as the profile representing the complainant without the protected attribute's effects on non-protected attributes, allowing the pair to differ in those attributes, in contrast to the ceteris paribus comparator that enforces an idealized difference only in the protected attribute. The mutatis mutandis comparator is presented as a more complex object whose implementation provides a venue for machine learning methods, illustrated through a real-world example that shows altered testing outcomes.

What carries the argument

The mutatis mutandis (MM) comparator, defined as the counterfactual profile of the complainant absent the protected attribute's effects on non-protected attributes.

If this is right

  • Discrimination testing tools that rely solely on ceteris paribus pairs may produce different evidence than tools using mutatis mutandis pairs.
  • Machine learning methods become necessary to estimate the adjustments required by the mutatis mutandis comparator.
  • Real-world discrimination tests can reach different conclusions once the mutatis mutandis comparator replaces the ceteris paribus one.
  • The causal modeling requirement applies to any protected attribute that influences non-protected attributes in the data.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • This framing could extend discrimination analysis to settings where protected attributes affect multiple downstream variables simultaneously.
  • Estimation of the mutatis mutandis comparator might benefit from existing causal inference methods for counterfactual prediction.
  • Automated decision systems could be audited by generating mutatis mutandis comparators for each complainant rather than fixed reference profiles.

Load-bearing premise

The counterfactual state of the complainant without the effects of the protected attribute on non-protected attributes can be meaningfully defined and estimated from data.

What would settle it

A dataset in which the mutatis mutandis comparator estimated from observed cases produces discrimination conclusions that contradict the conclusions from the ceteris paribus comparator, and independent verification of the actual decision process shows one set of conclusions is systematically wrong.

Figures

Figures reproduced from arXiv: 2405.13693 by Jose M. Alvarez, Salvatore Ruggieri.

Figure 1
Figure 1. Figure 1: The auxiliary causal knowledge with corresponding SCM [PITH_FULL_IMAGE:figures/full_fig_p015_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: We show the distributions in the form of box-plots for the [PITH_FULL_IMAGE:figures/full_fig_p018_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: The same analysis follows as in Figure 2 but based on the protected attribute race, including similar insights and conclusions [PITH_FULL_IMAGE:figures/full_fig_p018_3.png] view at source ↗
read the original abstract

Testing for individual discrimination involves deriving a profile, the comparator, similar to the one making the discrimination claim, the complainant, based on a protected attribute, such as race or gender, and comparing their decision outcomes. The complainant-comparator pair is central to discrimination testing. Most discrimination testing tools rely on this pair to establish evidence for discrimination. In this work, we revisit the role of the comparator in discrimination testing. We first argue for the inherent causal modeling nature of deriving the comparator. We then introduce a two-kind classification for the comparator: the ceteris paribus, or "with all else equal," (CP) comparator and the mutatis mutandis, or "with the appropriate adjustments being made," (MM) comparator. The CP comparator is the standard comparator, representing an idealized comparison for establishing discrimination as it aims for a complainant-comparator pair that only differs in membership in the protected attribute. As an alternative to the CP comparator, we define the MM comparator, which requires a comparator that represents the ``what would have been'' of the complainant without the effects of the protected attribute on the non-protected attributes. Under the MM comparator, the complainant-comparator pair can be dissimilar in terms of the non-protected attributes, departing from the idealized comparison imposed by the CP comparator. Notably, the MM comparator denotes a more complex object and its implementation offers an impactful venue for machine learning methods. We illustrate these two comparators and their impact on discrimination testing using a real-world example.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The paper argues that discrimination testing is inherently causal in nature and introduces a two-kind classification of the comparator: the ceteris paribus (CP) comparator, which seeks a pair differing only on the protected attribute, and the mutatis mutandis (MM) comparator, which represents the counterfactual state of the complainant in which the protected attribute has no effect on non-protected attributes (allowing dissimilarity on those attributes). It claims that the MM comparator is a more complex object whose implementation creates an impactful venue for machine learning methods, and illustrates the distinction with a real-world example.

Significance. If the MM comparator can be shown to be well-defined and identifiable, the distinction could shift discrimination testing away from idealized 'all else equal' comparisons toward more realistic counterfactuals, potentially creating new applications for causal ML techniques in fairness auditing and legal evidence.

major comments (2)
  1. [Abstract] Abstract and the definitions section: the central claim that the MM comparator 'denotes a more complex object and its implementation offers an impactful venue for machine learning methods' rests on the ability to define and estimate the required counterfactual state, yet no identification assumptions, structural causal model, or estimation procedure are supplied; without these the complexity and ML-venue assertions do not follow from the verbal definition alone.
  2. [real-world example] The real-world example section: the illustration compares CP and MM outcomes but supplies no quantitative assessment of how the MM counterfactual is constructed or validated against data, leaving the practical distinction between the two comparators untested.
minor comments (1)
  1. Notation for the two comparators remains verbal; introducing explicit mathematical notation (e.g., for the adjustment mapping in the MM case) would improve precision.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the thoughtful and constructive report. Our manuscript is a conceptual paper that introduces a causal distinction between two types of comparators in discrimination testing. We respond to each major comment below.

read point-by-point responses

  1. Referee: [Abstract] Abstract and the definitions section: the central claim that the MM comparator 'denotes a more complex object and its implementation offers an impactful venue for machine learning methods' rests on the ability to define and estimate the required counterfactual state, yet no identification assumptions, structural causal model, or estimation procedure are supplied; without these the complexity and ML-venue assertions do not follow from the verbal definition alone.

    Authors: We agree that realizing an operational MM comparator requires a structural causal model and identification assumptions. The manuscript's claim is narrower: the verbal definition of the MM comparator already shows it to be a more complex object than the CP comparator because it requires adjusting multiple non-protected attributes for the effect of the protected attribute. This complexity, by its nature, creates an opening for causal machine-learning techniques. We will add a brief discussion of possible identification strategies (e.g., under a linear structural causal model or using do-calculus) to make this link explicit. revision: partial

  2. Referee: [real-world example] The real-world example section: the illustration compares CP and MM outcomes but supplies no quantitative assessment of how the MM counterfactual is constructed or validated against data, leaving the practical distinction between the two comparators untested.

    Authors: The example is intended only to illustrate the conceptual difference in the resulting complainant-comparator pairs. We will expand the section to describe, at a high level, how the MM counterfactual could be approximated from observational data (e.g., via conditional expectation under an intervention on the protected attribute), while clarifying that a full quantitative construction and validation lies outside the scope of the present conceptual contribution. revision: partial

Circularity Check

0 steps flagged

No circularity: paper advances definitional distinction without reducing claims to inputs by construction

full rationale

The manuscript introduces CP and MM comparators via explicit definitions in the abstract and body, framing the MM version as requiring a counterfactual 'what would have been' state by verbal stipulation. No equations, fitted parameters, predictions, or self-citations are present in the supplied text that would make any central claim (e.g., MM complexity or ML venue) equivalent to its own inputs. The argument is self-contained as a conceptual reframing of discrimination testing, with the added complexity asserted directly from the definition rather than derived or fitted.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The paper introduces no free parameters, no new invented entities, and relies on standard causal modeling assumptions already present in the fairness literature.

axioms (1)
  • domain assumption Counterfactual states of non-protected attributes given absence of protected-attribute effects can be defined and estimated.
    Invoked when defining the MM comparator as the 'what would have been' state.

pith-pipeline@v0.9.0 · 5796 in / 1144 out tokens · 18379 ms · 2026-05-24T00:49:31.279459+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

107 extracted references · 107 canonical work pages · 1 internal anchor

  1. [1]

    Jeremias Adams-Prassl, Reuben Binns, and Aislinn Kelly-Lyth. 2022. Directly Discriminatory Algorithms.The Modern Law Review(2022)

    work page 2022

  2. [2]

    José M. Álvarez, Alejandra Bringas Colmenarejo, Alaa Elobaid, Simone Fabbrizzi, Miriam Fahimi, Antonio Ferrara, Siamak Ghodsi, Carlos Mougan, Ioanna Papageorgiou, Paula Reyero Lobo, Mayra Russo, Kristen M. Scott, Laura State, Xuan Zhao, and Salvatore Ruggieri. 2024. Policy advice and best practices on bias and fairness in AI.Ethics Inf. Technol.26, 2 (2024), 31

    work page 2024

  3. [3]

    Álvarez and Salvatore Ruggieri

    José M. Álvarez and Salvatore Ruggieri. 2023. Counterfactual Situation Testing: Uncovering Discrimination under Fairness given the Difference. In EAAMO. ACM, 2:1–2:11

    work page 2023

  4. [4]

    Álvarez and Salvatore Ruggieri

    José M. Álvarez and Salvatore Ruggieri. 2025. Counterfactual Situation Testing: From Single to Multidimensional Discrimination.J. Artif. Intell. Res.82 (2025), 2279–2323

    work page 2025

  5. [5]

    Jose M Alvarez and Salvatore Ruggieri. 2025. Toward A Causal Framework for Modeling Perception. InProceedings of the AAAI/ACM Conference on AI, Ethics, and Society, Vol. 8. 166–178

    work page 2025

  6. [6]

    2008.Mostly Harmless Econometrics

    Joshua D Angrist and Jörn-Steffen Pischke. 2008.Mostly Harmless Econometrics. Princeton University Press

    work page 2008

  7. [7]

    Jacy Reese Anthis and Victor Veitch. 2023. Causal Context Connects Counterfactual Fairness to Robust Prediction and Group Fairness. InNeurIPS

    work page 2023

  8. [8]

    Acting White

    David Austen-Smith and Roland G. Fryer. 2005. An Economic Analysis of “Acting White”.The Quarterly Journal of Economics120, 2 (2005), 551–583

    work page 2005

  9. [9]

    Rachel Baker, Daniel Klasik, and Sean F Reardon. 2018. Race and stratification in college enrollment over time.AERA Open4, 1 (2018)

    work page 2018

  10. [10]

    2019.Fairness and Machine Learning

    Solon Barocas, Moritz Hardt, and Arvind Narayanan. 2019.Fairness and Machine Learning. fairmlbook.org. http://www.fairmlbook.org

    work page 2019

  11. [11]

    Solon Barocas and Andrew D Selbst. 2016. Big data’s disparate impact.California Law Review104, 3 (2016), 671–732

    work page 2016

  12. [12]

    Marc Bendick. 2007. Situation Testing for Employment Discrimination in the United States of America.Horizons stratégiques3, 5 (2007), 17–39

    work page 2007

  13. [13]

    Yoshua Bengio, Li Yao, Guillaume Alain, and Pascal Vincent. 2013. Generalized Denoising Auto-Encoders as Generative Models. InNIPS. 899–907

    work page 2013

  14. [14]

    Marianne Bertrand and Esther Duflo. 2017. Field Experiments on Discrimination.Handbook of Economic Field Experiments1 (2017), 309–393

    work page 2017

  15. [15]

    Marianne Bertrand and Sendhil Mullainathan. 2004. Are Emily and Greg More Employable than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination.The American Economic Review94, 4 (2004), 991–1013

    work page 2004

  16. [16]

    Reuben Binns. 2020. On the apparent conflict between individual and group fairness. InFAT*. ACM, 514–524

    work page 2020

  17. [17]

    Emily Black, Samuel Yeom, and Matt Fredrikson. 2020. FlipTest: fairness testing via optimal transport. InFAT*. ACM, 111–121

    work page 2020

  18. [18]

    Eva Blanco-Mallo, Laura Morán-Fernández, Beatriz Remeseiro, and Verónica Bolón-Canedo. 2023. Do all roads lead to Rome? Studying distance measures in the context of machine learning.Pattern Recognit.141 (2023), 109646

    work page 2023

  19. [19]

    Eduardo Bonilla-Silva. 1997. Rethinking Racism: Toward a structural interpretation.American Sociological Review(1997), 465–480

    work page 1997

  20. [20]

    Mattia Cerrato, Marius Köppel, Roberto Esposito, and Stefan Kramer. 2023. Invariant Representations with Stochastically Quantized Neural Networks. InAAAI. AAAI Press, 6962–6970

    work page 2023

  21. [21]

    Mattia Cerrato, Marius Köppel, Alexander Segner, and Stefan Kramer. 2022. Fair Group-Shared Representations with Normalizing Flows.CoRR abs/2201.06336 (2022)

    work page arXiv 2022

  22. [22]

    Mattia Cerrato, Marius Köppel, Philipp Wolf, and Stefan Kramer. 2024. 10 Years of Fair Representations: Challenges and Opportunities.CoRR abs/2407.03834 (2024)

    work page arXiv 2024

  23. [23]

    Zhang, Max Hort, Mark Harman, and Federica Sarro

    Zhenpeng Chen, Jie M. Zhang, Max Hort, Mark Harman, and Federica Sarro. 2024. Fairness Testing: A Comprehensive Survey and Analysis of Trends.ACM Trans. Softw. Eng. Methodol.33, 5 (2024), 137:1–137:59. 26 Álvarez and Ruggieri

    work page 2024

  24. [24]

    Raj Chetty, Nathaniel Hendren, Maggie R Jones, and Sonya R Porter. 2020. Race and economic opportunity in the United States: An intergenerational perspective.The Quarterly Journal of Economics135, 2 (2020), 711–783

    work page 2020

  25. [25]

    Dylan Conger and Mark C Long. 2013. Gender gaps in college enrollment: The role of gender sorting across public high schools.Educational Researcher42, 7 (2013), 371–380

    work page 2013

  26. [26]

    Kimberlé Crenshaw. 1989. Demarginalizing the intersection of race and sex: A black feminist critique of antidiscrimination doctrine, feminist theory and antiracist politics.University of Chicago Legal Forum1989 (1989), 139–167. Issue 1

    work page 1989

  27. [27]

    2019.Invisible Women

    Caroline Criado-Perez. 2019.Invisible Women. Vintage

    work page 2019

  28. [28]

    Sanghamitra Dutta, Praveen Venkatesh, Piotr Mardziel, Anupam Datta, and Pulkit Grover. 2021. Fairness Under Feature Exemptions: Counterfactual and Observational Measures.IEEE Trans. Inf. Theory67, 10 (2021), 6675–6710

    work page 2021

  29. [29]

    Varshney

    Sanghamitra Dutta, Dennis Wei, Hazar Yueksel, Pin-Yu Chen, Sijia Liu, and Kush R. Varshney. 2020. Is There a Trade-Off Between Fairness and Accuracy? A Perspective Using Mismatched Hypothesis Testing. InICML (Proceedings of Machine Learning Research, Vol. 119). PMLR, 2803–2813

    work page 2020

  30. [30]

    Cynthia Dwork, Moritz Hardt, Toniann Pitassi, Omer Reingold, and Richard S. Zemel. 2012. Fairness through awareness. InITCS. ACM, 214–226

    work page 2012

  31. [31]

    Michael Fix and Raymond J. Struyk. 1993.Clear and Convincing Evidence: Measurement of Discrimination in America. Urban Institute Press

    work page 1993

  32. [32]

    European Union FRA. 2018. Handbook on European non-discrimination law. https://fra.europa.eu. Downloaded in 2023

    work page 2018

  33. [33]

    Roland G Fryer Jr and Paul Torelli. 2010. An empirical analysis of ‘acting white’.Journal of Public Economics94, 5-6 (2010), 380–396

    work page 2010

  34. [34]

    Lempitsky

    Yaroslav Ganin, Evgeniya Ustinova, Hana Ajakan, Pascal Germain, Hugo Larochelle, François Laviolette, Mario Marchand, and Victor S. Lempitsky

    work page

  35. [35]

    Domain-Adversarial Training of Neural Networks.J. Mach. Learn. Res.17 (2016), 59:1–59:35

    work page 2016

  36. [36]

    Claudia Goldin and Cecilia Rouse. 2000. Orchestrating Impartiality: The Impact of "Blind" Auditions on Female Musicians.American Economic Review90, 4 (September 2000), 715–741

    work page 2000

  37. [37]

    J. C. Gower. 1971. A General Coefficient of Similarity and Some of Its Properties.Biometrics27, 4 (1971), 857–871

    work page 1971

  38. [38]

    Riccardo Guidotti. 2024. Counterfactual explanations and how to find them: literature review and benchmarking.Data Min. Knowl. Discov.38, 5 (2024), 2770–2824

    work page 2024

  39. [39]

    Philipp Hacker. 2018. Teaching fairness to artificial intelligence: Existing and novel strategies against algorithmic discrimination under EU law. Common Market Law Review55, 4 (2018)

    work page 2018

  40. [40]

    Joseph Y. Halpern. 2016.Actual Causality. MIT Press

    work page 2016

  41. [41]

    Halpern and Judea Pearl

    Joseph Y. Halpern and Judea Pearl. 2001. Causes and Explanations: A Structural-Model Approach - Part I: Causes. InUAI. Morgan Kaufmann, 194–202

    work page 2001

  42. [42]

    Halpern and Judea Pearl

    Joseph Y. Halpern and Judea Pearl. 2001. Causes and Explanations: A Structural-Model Approach - Part II: Explanations. InIJCAI. Morgan Kaufmann, 27–34

    work page 2001

  43. [43]

    Alex Hanna, Emily Denton, Andrew Smart, and Jamila Smith-Loud. 2020. Towards a critical race methodology in algorithmic fairness. InFAT*. ACM, 501–512

    work page 2020

  44. [44]

    Moritz Hardt, Eric Price, and Nati Srebro. 2016. Equality of Opportunity in Supervised Learning. InNIPS. 3315–3323

    work page 2016

  45. [45]

    James J Heckman. 1998. Detecting Discrimination.Journal of Economic Perspectives12, 2 (1998), 101–116

    work page 1998

  46. [46]

    Lily Hu and Issa Kohler-Hausmann. 2020. What’s sex got to do with machine learning?. InFAT*. ACM, 513

    work page 2020

  47. [47]

    Ben Hutchinson and Margaret Mitchell. 2019. 50 Years of Test (Un)fairness: Lessons for Machine Learning. InFAT. ACM, 49–58

    work page 2019

  48. [48]

    Christina Ilvento. 2020. Metric Learning for Individual Fairness. InFORC (LIPIcs, Vol. 156). Schloss Dagstuhl - Leibniz-Zentrum für Informatik, 2:1–2:11

    work page 2020

  49. [49]

    Adrián Javaloy, Pablo Sánchez-Martín, and Isabel Valera. 2023. Causal normalizing flows: from theory to practice. InNeurIPS

    work page 2023

  50. [50]

    Faisal Kamiran and Toon Calders. 2009. Classifying without discriminating. In2009 2nd international conference on computer, control and communication. IEEE, 1–6

    work page 2009

  51. [51]

    Faisal Kamiran, Indre Zliobaite, and Toon Calders. 2013. Quantifying explainable discrimination and removing illegal discrimination in automated decision making.Knowl. Inf. Syst.35, 3 (2013), 613–644

    work page 2013

  52. [52]

    Niki Kilbertus, Mateo Rojas-Carulla, Giambattista Parascandolo, Moritz Hardt, Dominik Janzing, and Bernhard Schölkopf. 2017. Avoiding Discrimination through Causal Reasoning. InNIPS. 656–666

    work page 2017

  53. [53]

    Kingma and Max Welling

    Diederik P. Kingma and Max Welling. 2014. Auto-Encoding Variational Bayes. InICLR

    work page 2014

  54. [54]

    Discrimination in the Age of Algorithms

    Jon M. Kleinberg, Jens Ludwig, Sendhil Mullainathan, and Cass R. Sunstein. 2019. Discrimination in the Age of Algorithms.CoRRabs/1902.03731 (2019)

    work page internal anchor Pith review Pith/arXiv arXiv 2019

  55. [55]

    Issa Kohler-Hausmann. 2019. Eddie Murphy and the Dangers of Counterfactual Causal Thinking About Detecting Racial Discrimination. Northwestern University Law Review113, 5 (2019), 1163

    work page 2019

  56. [56]

    Adriano Koshiyama, Emre Kazim, Philip Treleaven, Pete Rai, Lukasz Szpruch, Giles Pavey, Ghazi Ahamat, Franziska Leutner, Randy Goebel, Andrew Knight, Janet Adams, Christina Hitrova, Jeremy Barnett, Parashkev Nachev, David Barber, Tomas Chamorro-Premuzic, Konstantin Klemmer, Miro Gregorovic, Shakeel Khan, Elizabeth Lomas, Airlie Hilliard, and Siddhant Chat...

    work page 2024

  57. [57]

    Brian Kulis. 2013. Metric Learning: A Survey.Found. Trends Mach. Learn.5, 4 (2013), 287–364

    work page 2013

  58. [58]

    Kusner, Joshua R

    Matt J. Kusner, Joshua R. Loftus, Chris Russell, and Ricardo Silva. 2017. Counterfactual Fairness. InNIPS. 4066–4076

    work page 2017

  59. [59]

    Kasper Lippert-Rasmussen. 2006. The badness of discrimination.Ethical Theory and Moral Practice9 (2006), 167–185. Mutatis Mutandis 27

    work page 2006

  60. [60]

    Joshua R. Loftus. 2024. Position: The Causal Revolution Needs Scientific Pragmatism. InICML. OpenReview.net

    work page 2024

  61. [61]

    Michele Loi, Francesco Nappo, and Eleonora Viganò. 2023. How I Would have been Differently Treated. Discrimination Through the Lens of Counterfactual Fairness.Res Publica29, 2 (2023), 185–211

    work page 2023

  62. [62]

    Christos Louizos, Kevin Swersky, Yujia Li, Max Welling, and Richard S. Zemel. 2016. The Variational Fair Autoencoder. InICLR

    work page 2016

  63. [63]

    David Madras, Elliot Creager, Toniann Pitassi, and Richard S. Zemel. 2018. Learning Adversarially Fair and Transferable Representations. InICML (Proceedings of Machine Learning Research, Vol. 80). PMLR, 3381–3390

    work page 2018

  64. [64]

    Subha Maity, Debarghya Mukherjee, Mikhail Yurochkin, and Yuekai Sun. 2020. There is no trade-off: enforcing fairness can improve accuracy. CoRRabs/2011.03173 (2020)

    work page arXiv 2020

  65. [65]

    Karima Makhlouf, Sami Zhioua, and Catuscia Palamidessi. 2024. When causality meets fairness: A survey.J. Log. Algebraic Methods Program.141 (2024), 101000

    work page 2024

  66. [66]

    Ron Mallon. 2007. A field guide to social construction.Philosophy Compass2, 1 (2007), 93–108

    work page 2007

  67. [67]

    2018.Statistical rethinking: A Bayesian course with examples in R and Stan

    Richard McElreath. 2018.Statistical rethinking: A Bayesian course with examples in R and Stan. Chapman and Hall/CRC

    work page 2018

  68. [68]

    Ninareh Mehrabi, Fred Morstatter, Nripsuta Saxena, Kristina Lerman, and Aram Galstyan. 2022. A Survey on Bias and Fairness in Machine Learning.ACM Comput. Surv.54, 6 (2022), 115:1–115:35

    work page 2022

  69. [69]

    Benn Michaels and Adolph Reed. 2020. The Trouble with Disparity. https://nonsite.org/the-trouble-with-disparity/. Nonsite, Issue num. 32

    work page 2020

  70. [70]

    Tim Miller. 2019. Explanation in artificial intelligence: Insights from the social sciences.Artif. Intell.267 (2019), 1–38

    work page 2019

  71. [71]

    Allison C Morgan, Samuel F Way, Michael JD Hoefer, Daniel B Larremore, Mirta Galesic, and Aaron Clauset. 2021. The Unequal Impact of Parenthood in Academia.Science Advances7, 9 (2021)

    work page 2021

  72. [72]

    Daniel Moyer, Shuyang Gao, Rob Brekelmans, Aram Galstyan, and Greg Ver Steeg. 2018. Invariant Representations without Adversarial Training. InNeurIPS. 9102–9111

    work page 2018

  73. [73]

    Ana Rita Nogueira, Andrea Pugnana, Salvatore Ruggieri, Dino Pedreschi, and João Gama. 2022. Methods and tools for causal discovery and causal inference.WIREs Data Mining Knowl. Discov.12, 2 (2022)

    work page 2022

  74. [74]

    Eirini Ntoutsi, Pavlos Fafalios, Ujwal Gadiraju, Vasileios Iosifidis, Wolfgang Nejdl, Maria-Esther Vidal, Salvatore Ruggieri, Franco Turini, Symeon Papadopoulos, Emmanouil Krasanakis, Ioannis Kompatsiaris, Katharina Kinder-Kurlanda, Claudia Wagner, Fariba Karimi, Miriam Fernández, Harith Alani, Bettina Berendt, Tina Kruegel, Christian Heinze, Klaus Broele...

    work page 2020

  75. [75]

    Cecilia Panigutti, Ronan Hamon, Isabelle Hupont, David Fernández Llorca, Delia Fano Yela, Henrik Junklewitz, Salvatore Scalzo, Gabriele Mazzini, Ignacio Sanchez, Josep Soler Garrido, and Emilia Gómez. 2023. The role of explainable AI in the context of the AI Act. InFAccT. ACM, 1139–1150

    work page 2023

  76. [76]

    2009.Causality(2 ed.)

    Judea Pearl. 2009.Causality(2 ed.). Cambridge University Press

    work page 2009

  77. [77]

    Judea Pearl, Madelyn Glymour, and Nicholas P. Jewell. 2016.Causal Inference in Statistics: A Primer. John Wiley & Sons

    work page 2016

  78. [78]

    Dino Pedreschi, Salvatore Ruggieri, and Franco Turini. 2008. Discrimination-aware data mining. InKDD. ACM, 560–568

    work page 2008

  79. [79]

    Drago Plecko and Elias Bareinboim. 2024. Causal Fairness Analysis: A Causal Toolkit for Fair Machine Learning.Found. Trends Mach. Learn.17, 3 (2024), 304–589

    work page 2024

  80. [80]

    Bilal Qureshi, Faisal Kamiran, Asim Karim, Salvatore Ruggieri, and Dino Pedreschi. 2020. Causal inference for social discrimination reasoning.J. Intell. Inf. Syst.54, 2 (2020), 425–437

    work page 2020

Showing first 80 references.