Data construction can embed social values and patterns of privilege into data itself, perpetuating testimonial injustice¶

Claim · All claims
Evidence · 1 study (2 entries) · theoretical r2–r3 · q1
Currency · newest study 2021 · not yet checked against its sources · not reviewed by a person · overdue for a recheck since 2024: tool and model accuracy changes quickly
Evidence quality · No study behind this claim is a peer-reviewed article with a DOI: it rests on a source with no registered publication record (a web page, book or unpublished report). It is kept because 2 design pages cite it; treat it as a lead to check, not an established finding.

Subclaims¶

q1 i? The process of constructing data builds social values and privilege into the data, so datafication cannot transform bad inputs into good outputs. → Haarman 2021 q1 i? Algorithms cannot perform the reflection upon and revision of prejudiced attitudes over time that Fricker held necessary to repair testimonial injustice, since they are predictive and forward-looking. → Haarman 2021 (2)

Evidence¶

Haarman 2021¶

Haarman, Susan. (2021). The Data Should Not Speak for Itself: Epistemic Injustice and Data as Rhetoric. Philosophical Studies in Education 52. https://www.ovpes.org/

q1 · i? · theoretical · r2

Theoretical argument on exclusivity: certain testimonies are left out of data by the collection process, and marginalized groups are less likely to be part of data producing actions. The article quotes Johnson that "'Garbage in, garbage out' is a central concept in data ethics."

"Regardless of whether this programmed bias was malicious or a result of ignorance, the process of constructing data builds social values and patterns of privilege into the data."

Haarman 2021 (2)¶

Haarman, Susan. (2021). The Data Should Not Speak for Itself: Epistemic Injustice and Data as Rhetoric. Philosophical Studies in Education 52. https://www.ovpes.org/

q1 · i? · theoretical · r3

The article argues algorithms may dismiss contradicting data if it is in the minority, and that more data can produce overfitting and false confidence rather than surfacing suppressed testimony.

"Additionally algorithms cannot participate in what Fricker said was necessary to prevent or repair the harm done by testimonial injustice – reflection upon and revision of prejudiced attitudes over time. They are predictive and forward-looking."

Discussion¶

Related Claims¶

• Big Data can amplify suppressed testimony because its sheer size lends it assumed authenticity — reports the opposite
• Big Data's variety can aid redress of hermeneutical injustice by letting buried experiences find themselves within the data — related
• Public accountability pressure on school-level improvement data can suppress honest inquiry and incentivize superficial fixes — related
• Machine learning systems can perpetuate biases contained in their training data — related
• Algorithmic bias and data privacy are central ethical risks of AI in adult education — related
• In a pedagogical space, epistemic injustice might be activated when teachers' identity prejudice leads them to give a student less credit, and the curriculum can block or enable students' understanding and expression of their social experiences — related
• In the analyzed school ethnography, stereotyping operates as a rationalization process that constructs uniformities, and the ghetto label privileges white culture's standards and evaluation networks — related
• Cultural processes and forms of epistemic injustice should be taken into account in understanding how inequalities in education are produced and maintained — a broader claim this one bears on