Pith. sign in

REVIEW 4 major objections 6 minor 98 references

What's Privacy Good for? Measuring Privacy as a Shield from Harms due to Personal Data Use

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper argues that privacy is best understood as a shield against harms from personal data use and demonstrates that a 14-item survey can measure it.

desk verdict A usable 14-item privacy-harm scale with strong reliability, but the validity claim needs discriminant evidence before it can stand. read the letter →

arxiv 2506.22787 v1 pith:OPCKSGDM submitted 2025-06-28 cs.CR cs.CY

Sri Harsha Gajavalli , Junichi Koizumi , Rakibul Hasan This is my paper
classification cs.CRcs.CY
keywords privacyharmharm-centricperceivedmeasurementAIdecision-makingeducationdataemploymentalgorithmicfairnesscontextualintegrity
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper argues that privacy is best understood not as hiding data or controlling its flow, but as a shield against the harms that can arise from how personal data is used. To test that idea, the authors surveyed 400 college and university students who rated whether an AI using each of six inferred attributes could cause each of 14 harms in two consequential decisions: predicting course dropout and recommending job candidates. The ratings cohered statistically into a single underlying construct the authors call privacy harm, and the same data types produced different harms in the education and employment contexts. If the claim holds, a 14-item scale gives researchers and regulators a practical way to measure privacy in AI decision-making and to aim protections at the harms people actually fear.

What carries the argument

The load-bearing object is a 14-item perceived-privacy-harm scale, with items such as "Use of [attribute] to decide [dropout risk or job fit] can lead to biased decisions" rated on a five-point Likert scale. Each of the 14 harms is paired with each of six inferred data attributes, and the scale is administered in either an education or an employment scenario. The argument that the items measure one construct rests on two statistical tools: Cronbach's alpha for internal consistency and factor analysis with scree plots showing a single dominant factor. This machinery turns the philosophical claim that privacy is a shield against harms into a directly testable measurement instrument.

What would settle it

Run the same 14-item instrument alongside a behavioral choice, for example letting participants opt whether an AI uses each inferred attribute in a real or simulated dropout or hiring decision, and check whether harm ratings predict refusal or experienced harm; if they do not track actual behavior or outcomes, the scale measures anticipated discomfort rather than privacy harm.

Watch

Extended reading notes

Core claim

The paper's central claim is that measuring perceived harms from personal data use is a reliable and valid quantification of people's conceptualization of privacy as a shield against such harms. In an online study, 400 students rated statements of the form "Use of [attribute] to decide [dropout risk or job fit] can lead to [harm]" for six inferred attributes—demographics, personality traits, emotional state, motivation, creativity and problem solving, and physical or cognitive impairment—across 14 harms including bias, stereotyping, manipulation, wrong inference, loss of autonomy, and stress. Cronbach's alpha values between 0.93 and 0.96, leave-one-out item analyses, and factor analyses showing one dominant factor support the authors' conclusion that the 14 items are internally consistent and represent a general notion of privacy harm. Participants anticipated harms even for data they did not consider inherently private, the same attribute produced different harms in education versus employment, and demographic subgroups differed in which harms they feared. The authors further mapped the 14 harm categories onto 35 of 36 education- and employment-related incidents in the AI Incident Database and take this as evidence of external validity.

Load-bearing premise

The entire empirical case rests on treating participants' ratings of hypothetical survey scenarios as accurate reports of the harms they would actually experience from AI data use, with no behavioral or objective-harm measure included to confirm that link.

Editorial extensions

If this is right

  • If the scale measures what it claims, privacy researchers can replace abstract 'concern' questions with concrete harm ratings in any new application context.
  • The finding that non-private data still generate perceived harms implies that privacy protections cannot be limited to sensitive categories such as demographics or disability.
  • Because the same attribute produced different harms in education and employment, interventions should be context-specific rather than attribute-specific.
  • Heterogeneous harm perceptions across gender, age, race, and education level imply that equitable privacy policy should prioritize the harms reported by the most affected subgroups, not only majority norms.
  • The framework directly supports threat modeling: a system can be audited by asking which of the 14 harms each data use could cause, and technical fixes such as adversarial censoring can target those harms.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural next test is whether harm ratings predict actual protective behavior, such as refusing to share data when given a real choice; the current study measures only stated agreement with hypothetical scenarios.
  • Because all 14 items loaded on one factor, the scale may be capturing a broader risk perception or algorithm aversion rather than distinct harm types, so discriminant-validity checks against unrelated technology attitudes would clarify the construct.
  • The same instrument could be carried into adjacent decision contexts such as credit scoring, health triage, or content moderation, with specific predictions about which harms dominate when inferred attributes drive consequential choices.
  • The AI Incident Database mapping was performed by the lead author and reviewed by a second author; independent multi-rater coding of a larger incident sample would test how comprehensively the taxonomy covers real-world cases.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes a harm-centric conceptualization of privacy, defining privacy as a shield against harms arising from personal data use, and operationalizes it with a 14-item Likert instrument administered to 400 college/university students across two decision contexts (AI-based dropout prediction and AI-based hiring). The authors report high internal consistency (Cronbach's alpha 0.93–0.96, with item-total correlations), a single-factor structure from scree plots, differential harm perceptions across data types and demographic groups, and an external mapping of the 14 harm categories onto 35 AI Incident Database entries. The central claim is that measuring perceived harms from personal data use is a reliable and valid quantification of people's conceptualization of privacy as a shield against such harms.

Significance. If the validity claim were fully supported, the paper would make an important contribution: a compact, reliability-tested instrument for measuring perceived privacy harm in AI decision contexts, a useful complement to frameworks like Contextual Integrity that struggle with inferred data and novel harms. The paper also provides a credible demonstration of high internal consistency, a thoughtful procedural-justice framing, and a taxonomy-mapping exercise that shows the 14 harm categories are descriptive of real-world AI incidents. The demographic variation results, if confirmed, would support the paper's equity-oriented recommendations. However, the empirical evidence currently establishes reliability and descriptive breadth more strongly than it establishes measurement validity, and the central 'valid' inference requires additional psychometric evidence.

major comments (4)
  1. [§4.2] The claim that the 14 items 'collectively represent a general notion of privacy harms' is supported only by scree plots, total variance explained, and Cronbach's alpha. No factor loadings, parallel analysis, or comparison with alternative factor models are reported, and all 14 items are negatively valenced statements about AI decisions (e.g., 'can lead to biased decisions', 'will create stress for me', 'violates my privacy'). A single general factor could therefore reflect negative affect toward AI or a general evaluative response tendency rather than a privacy-specific construct. No discriminant validity evidence against general privacy concern scales, AI attitudes, or negative affectivity is provided. Since the abstract and conclusion hinge on the word 'valid', this gap is load-bearing and needs to be closed with convergent and discriminant validity analyses.
  2. [§4.5] The AI Incident Database coding exercise demonstrates that the 14 harm categories can be applied to real-world incidents, but it does not validate the survey responses as a measure of people's privacy conceptualization. Coding incident descriptions is an expert taxonomy-mapping task, not a test of whether the Likert scores predict or correlate with actual privacy harms, protective behavior, or other criterion variables. Additionally, the coding was done primarily by the lead author with no reported inter-rater reliability statistic. The section should be repositioned as descriptive applicability of the taxonomy, not as evidence for the measurement validity of the scale.
  3. [Appendix A.1.2] The questionnaire appendix states that participants were asked about 'the three attributes' and provides descriptions for only Demographics, Personality Traits, and Emotional State, while the method and results sections report six data types: Demographics, Personality Traits, Emotional State, Motivation, Creativity and Problem Solving, and Physical or Cognitive Impairment. Descriptions and survey wording for the latter three attributes are missing, making the study irreproducible and leaving ambiguity about whether each participant rated all six attributes or a subset. This discrepancy must be corrected in a revision.
  4. [§4.4.1 and §4.4.2] The text states that demographic comparisons are reported 'after applying Bonferroni correction', but many listed p-values (e.g., p = 0.046, p = 0.03, p = 0.02, p = 0.01) would not survive a standard Bonferroni correction for the number of comparisons made across 14 harms, 6 data types, and multiple demographic groupings. The correction procedure is not described, and the reported p-values are inconsistent with the stated correction. This undermines the reliability of the demographic-difference findings and needs to be resolved, either by reporting adjusted p-values or by describing the actual multiple-comparison procedure used.
minor comments (6)
  1. [Author affiliations] 'School of Coumputing and Augmented Intelligence' contains a typo; it should read 'Computing'.
  2. [§2.1.3] The manuscript refers to 'Jacabi et al.' when discussing the taxonomy of privacy risks; the cited work appears to be by Jakobi et al., and the name should be corrected throughout.
  3. [Figure 2 caption] The caption says 'Mean and SD of perceived harms', but the text describes standard error bars; please align the caption with the plotted statistic.
  4. [§4.5] The sentence 'The successful application of the harm categories to label all (except one) of the incidents the validity of our harm taxonomy' is missing a verb (e.g., 'demonstrates') and should be rewritten.
  5. [§4.4.1] There are inconsistent notation and formatting issues in the reported statistics, such as 'M eanf', 'MGrad+', and 'MPG' being used interchangeably; these should be unified.
  6. [§3.2] The power analysis based on 20 pilot participants is mentioned without specifying the target test, effect size, or software used; adding these details would improve reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reliability and factor-analytic results are empirical, the AIID mapping is an external benchmark, and the self-citations are not load-bearing.

full rationale

The paper's central claim is that perceived harms from personal data use can reliably and validly quantify a harm-centric conception of privacy. The reliability evidence (Cronbach's alpha 0.93-0.96, item-total correlations, leave-one-out analyses in Section 4.2) is an empirical property of the 14 survey items; the items could have failed to cohere, so the finding is not true by construction. The factor analysis (scree plots, 52-66% variance explained) is likewise an empirical dimensionality result; labelling the single factor 'privacy harm' is an interpretive step, not a derivation from the item wording. The AIID coding in Section 4.5 uses the 14 harm categories to label 35 real-world incidents; this is an external corpus independent of the survey responses, and while it supports the taxonomy's descriptive applicability rather than criterion validity of the Likert scores, that is a limitation, not circularity. The two self-citations ([52] Hasan and Fritz; [97] Kelso et al.) appear only in background and recommendation contexts and are not load-bearing for the measurement claim. The construct-validity gap - no discriminant validation against general AI attitudes or negative affectivity - is a substantive correctness and validation concern, but it does not make the derivation circular under the standards applied here.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The paper introduces no new physical or mathematical entities or fitted parameters. Its burden is in domain assumptions: self-report as proxy for harm, representative item selection, statistical criteria for unidimensionality, sample generalizability, and subjective coding as external validation. These are acknowledged in the paper's limitations section to varying degrees.

assumptions (5)
  • domain assumption Participants' Likert agreement with harm statements measures the actual privacy harms they would experience from AI-based decisions.
    Invoked throughout §3.2 and §4; all empirical conclusions derive from self-reported perceptions of hypothetical scenarios, with no objective or behavioral validation.
  • domain assumption The six selected data types and 14 harm types are representative of personal-data-driven AI harms in education and employment.
    Selection in §3.1.2 and §3.1.3 is based on literature review; no coverage or saturation analysis supports completeness.
  • domain assumption Cronbach's alpha and scree-plot elbow are sufficient to establish that a single latent factor explains the 14 items.
    §4.2 uses these without confirmatory factor analysis, parallel analysis, or reported factor loadings.
  • domain assumption College students recruited via Prolific are 'experiential experts' whose perceptions generalize to the broader student and job-seeker population.
    §3.1.1 justifies the sample; §3.2 acknowledges sampling bias but the generalizability claim remains.
  • domain assumption The lead author's coding of AI Incident Database incidents is a valid external benchmark for the harm taxonomy.
    §4.5 relies on subjective incident tagging by one author with a second reviewer; no inter-rater reliability metric is reported.

how reviews work

0 comments
Cite this review

Pith. "Pith review of What's Privacy Good for? Measuring Privacy as a Shield from Harms due to Personal Data Use." pith.science (2026). https://pith.science/paper/OPCKSGDM

@misc{pith2026250622787,
  author       = {Pith},
  title        = {Pith review of: What's Privacy Good for? Measuring Privacy as a Shield from Harms due to Personal Data Use},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OPCKSGDM}},
  note         = {Machine review of arXiv:2506.22787}
}
read the original abstract

We propose a harm-centric conceptualization of privacy that asks: What harms from personal data use can privacy prevent? The motivation behind this research is limitations in existing privacy frameworks (e.g., Contextual Integrity) to capture or categorize many of the harms that arise from modern technology's use of personal data. We operationalize this conceptualization in an online study with 400 college and university students. Study participants indicated their perceptions of different harms (e.g., manipulation, discrimination, and harassment) that may arise when artificial intelligence-based algorithms infer personal data (e.g., demographics, personality traits, and cognitive disability) and use it to identify students who are likely to drop out of a course or the best job candidate. The study includes 14 harms and six types of personal data selected based on an extensive literature review. Comprehensive statistical analyses of the study data show that the 14 harms are internally consistent and collectively represent a general notion of privacy harms. The study data also surfaces nuanced perceptions of harms, both across the contexts and participants' demographic factors. Based on these results, we discuss how privacy can be improved equitably. Thus, this research not only contributes to enhancing the understanding of privacy as a concept but also provides practical guidance to improve privacy in the context of education and employment.

Figures

Figures reproduced from arXiv: 2506.22787 by the authors.

Figure 1
Figure 1. Scree plot for factor analysis [PITH_FULL_IMAGE:figures/full_fig_p009_1.png] view at source ↗
Figure 2
Figure 2. Mean and SD of perceived harms in education and employment. [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗
Figure 3
Figure 3. Distribution of participants’ responses to harm statements (left: education, right: employment). [PITH_FULL_IMAGE:figures/full_fig_p023_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

98 extracted references · 75 canonical work pages

  1. [1]

    Smith, and Paul Hart

    Heng Xu, Tamara Dinev, H. Smith, and Paul Hart. Examining the Formation of Individual’s Privacy Concerns: Toward an Integrative View.ICIS 2008 Proceedings, January 2008

    2008

  2. [2]

    Daniel J. Solove. Data Is What Data Does: Regulating Based on Harm and Risk Instead of Sensitive Data, January 2024. 17 A PREPRINT - S EPTEMBER 18, 2025 Table 4: Number of participants when grouped by demographic factors Grouping Education context Employment context White 112 107 Non-White 75 82 STEM 90 99 Non-STEM 110 101 Male 89 84 Female 97 102 Undergr...

    2024

  3. [3]

    An Extended Privacy Calculus Model for E-Commerce Transactions

    Tamara Dinev and Paul Hart. An Extended Privacy Calculus Model for E-Commerce Transactions. Information Systems Research, 17(1):61–80, March 2006. Publisher: INFORMS

    2006

  4. [4]

    Daniel J. Solove. The Myth of the Privacy Paradox. George Washington Law Review, 89:1, 2021

    2021

  5. [5]

    Daniel J. Solove. Murky Consent: An Approach to the Fictions of Consent in Privacy Law. SSRN Electronic Journal, 2023

    2023

  6. [6]

    Privacy as Contextual Integrity

    Helen Nissenbaum. Privacy as Contextual Integrity. 79:119, 2004. Publisher: HeinOnline

    2004

  7. [7]

    Kumar, Michael Zimmer, and Jessica Vitak

    Priya C. Kumar, Michael Zimmer, and Jessica Vitak. A Roadmap for Applying the Contextual Integrity Framework in Qualitative Privacy Research. Proc. ACM Hum.-Comput. Interact., 8(CSCW1):219:1–219:29, April 2024

    2024

  8. [8]

    Distinguishing privacy law

    Marı’a P Angel and Ryan Calo. Distinguishing privacy law. Columbia Law Review, 124(2):507–562, 2024. Publisher: JSTOR

    2024

Show all 98 references
  1. [9]

    Richards

    Neil M. Richards. Why Privacy Matters: An Introduction, December 2021

    2021

  2. [10]

    Privacy harm exceptionalism

    Ryan Calo. Privacy harm exceptionalism. 12:361, 2014. Publisher: HeinOnline

    2014

  3. [11]

    Jaspreet Bhatia and Travis D. Breaux. Empirical Measurement of Perceived Privacy Risk. ACM Trans. Comput.- Hum. Interact., 25(6):34:1–34:47, December 2018

    2018

  4. [12]

    Is AI recruiting (un)ethical? A human rights perspective on the use of AI for hiring

    Anna Lena Hunkenschroer and Alexander Kriebitz. Is AI recruiting (un)ethical? A human rights perspective on the use of AI for hiring. AI and Ethics, 3(1):199–213, February 2023

    2023

  5. [13]

    Understanding dropouts in MOOCs

    Wenzheng Feng, Jie Tang, and Tracy Xiao Liu. Understanding dropouts in MOOCs. Proceedings of the AAAI Conference on Artificial Intelligence, 33(01):517–524, July 2019

    2019

  6. [14]

    OULAD MOOC dropout and result prediction using ensemble, deep learning and regression techniques

    Nikhil Indrashekhar Jha, Ioana Ghergulescu, and Arghir-Nicolae Moldovan. OULAD MOOC dropout and result prediction using ensemble, deep learning and regression techniques. In CSEDU (2), pages 154–164, 2019

    2019

  7. [15]

    Predictive analytic models of student success in higher education: A review of methodology

    Ying Cui, Fu Chen, Ali Shiri, and Yaqin Fan. Predictive analytic models of student success in higher education: A review of methodology. Information and Learning Sciences, 2019. Publisher: Emerald Publishing Limited

    2019

  8. [16]

    ResearchGate, October 2024

    (PDF) The Impact of AI on Recruitment and Selection Processes: Analysing the role of AI in automating and enhancing recruitment and selection procedures. ResearchGate, October 2024

    2024

  9. [17]

    Predictive Analytics for Employee Retention: Forecasting and Preventing Turnover - Hirebee

  10. [18]

    An overview of ethical issues in using AI systems in hiring with a case study of Amazon’s AI based hiring tool

    Akhil Alfons Kodiyan. An overview of ethical issues in using AI systems in hiring with a case study of Amazon’s AI based hiring tool. Researchgate Preprint, pages 1–19, 2019

    2019

  11. [19]

    Toward inclusive tech policy design: a method for underrep- resented voices to strengthen tech policy documents

    Meg Young, Lassana Magassa, and Batya Friedman. Toward inclusive tech policy design: a method for underrep- resented voices to strengthen tech policy documents. Ethics and Information Technology, 21(2):89–103, June 2019

    2019

  12. [20]

    Knijnenburg, Xinru Page, Pamela Wisniewski, Heather Richter Lipford, Nicholas Proferes, and Jennifer Romano, editors

    Bart P. Knijnenburg, Xinru Page, Pamela Wisniewski, Heather Richter Lipford, Nicholas Proferes, and Jennifer Romano, editors. Modern Socio-Technical Perspectives on Privacy. Springer International Publishing, Cham, 2022

    2022

  13. [21]

    Un-Paradoxing Privacy: Considering Hopeful Trust

    Bran Knowles and Stacey Conchie. Un-Paradoxing Privacy: Considering Hopeful Trust. ACM Trans. Comput.- Hum. Interact., 30(6):87:1–87:24, September 2023. 18 A PREPRINT - S EPTEMBER 18, 2025

    2023

  14. [22]

    Mhaidli, Kentaro Toyama, and Florian Schaub

    Tamy Guberek, Allison McDonald, Sylvia Simioni, Abraham H. Mhaidli, Kentaro Toyama, and Florian Schaub. Keeping a Low Profile? Technology, Risk and Privacy among Undocumented Immigrants. In Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems, CHI ’18, ...

    2018

  15. [23]

    Unpacking "privacy" for a networked world

    Leysia Palen and Paul Dourish. Unpacking "privacy" for a networked world. In Proceedings of the SIGCHI Conference on Human Factors in Computing Systems, CHI ’03, pages 129–136, New York, NY , USA, April 2003. Association for Computing Machinery

    2003

  16. [24]

    Taking Trust Seriously in Privacy Law

    Neil Richards and Woodrow Hartzog. Taking Trust Seriously in Privacy Law. Stanford Technology Law Review, 19:431, 2015

    2015

  17. [25]

    Successful failure: what Foucault can teach us about privacy self-management in a world of Facebook and big data

    Gordon Hull. Successful failure: what Foucault can teach us about privacy self-management in a world of Facebook and big data. Ethics and Information Technology, 17(2):89–101, June 2015

    2015

  18. [26]

    Solove and Woodrow Hartzog

    Daniel J. Solove and Woodrow Hartzog. Kafka in the Age of AI and the Futility of Privacy as Control, January 2024

    2024

  19. [27]

    Privacy in context: Technology, policy, and the integrity of social life

    Helen Nissenbaum. Privacy in context: Technology, policy, and the integrity of social life. Stanford University Press, 2009

    2009

  20. [28]

    Contextual Integrity Up and Down the Data Food Chain

    Helen Nissenbaum. Contextual Integrity Up and Down the Data Food Chain. Theoretical Inquiries in Law , 20(1):221–256, January 2019. Publisher: De Gruyter

    2019

  21. [29]

    Berryessa

    Rebecca Umbach, Nicola Henry, Gemma Faye Beard, and Colleen M. Berryessa. Non-Consensual Synthetic Intimate Imagery: Prevalence, Attitudes, and Knowledge in 10 Countries. In Proceedings of the CHI Conference on Human Factors in Computing Systems, CHI ’24, pages 1–20, New York,...

    2024

  22. [30]

    Socially Disruptive Technologies, Contextual Integrity, and Conservatism About Moral Change

    Ibo van de Poel. Socially Disruptive Technologies, Contextual Integrity, and Conservatism About Moral Change. Philosophy & Technology, 35(3):82, August 2022

    2022

  23. [31]

    The meaning and value of privacy

    Daniel J Solove. The meaning and value of privacy. Social dimensions of privacy: Interdisciplinary perspectives, 71, 2015. Publisher: Cambridge University Press Cambridge

    2015

  24. [32]

    Investigating people’s privacy risk perception.Proc

    Nina Gerber, Benjamin Reinheimer, and Melanie V olkamer. Investigating people’s privacy risk perception.Proc. Priv. Enhancing Technol., 2019(3):267–288, 2019

    2019

  25. [33]

    It Is About What They Could Do with the Data: A User Perspective on Privacy in Smart Metering

    Timo Jakobi, Sameer Patil, Dave Randall, Gunnar Stevens, and V olker Wulf. It Is About What They Could Do with the Data: A User Perspective on Privacy in Smart Metering. ACM Trans. Comput.-Hum. Interact., 26(1):2:1–2:44, January 2019

    2019

  26. [34]

    Is There a Reverse Privacy Paradox? An Exploratory Analysis of Gaps Between Privacy Perspectives and Privacy-Seeking Behaviors, July 2023

    Jessica Colnago, Lorrie Faith Cranor, and Alessandro Acquisti. Is There a Reverse Privacy Paradox? An Exploratory Analysis of Gaps Between Privacy Perspectives and Privacy-Seeking Behaviors, July 2023

    2023

  27. [35]

    YES, FIRMS HA VE MY DATA BUT WHAT DOES IT MATTER? MEASURING PRIV ACY RISKS.Research Papers, November 2018

    Sabrina Karwatzki, Manuel Trenz, and Daniel Veit. YES, FIRMS HA VE MY DATA BUT WHAT DOES IT MATTER? MEASURING PRIV ACY RISKS.Research Papers, November 2018

    2018

  28. [36]

    A Taxonomy of user-perceived privacy risks to foster accountability of data-based services

    Timo Jakobi, Maximilian von Grafenstein, Patrick Smieskol, and Gunnar Stevens. A Taxonomy of user-perceived privacy risks to foster accountability of data-based services. Journal of Responsible Technology, 10:100029, July 2022

    2022

  29. [37]

    Daniel J. Solove. A Taxonomy of Privacy. University of Pennsylvania Law Review, 154(3):477–564, 2005

    2005

  30. [38]

    An Introduction to Privacy Engineering and Risk Management in Federal Systems

    Sean Brooks, Michael Garcia, Naomi Lefkovitz, Suzanne Lightman, and Ellen Nadeau. An Introduction to Privacy Engineering and Risk Management in Federal Systems. Technical Report NIST Internal or Interagency Report (NISTIR) 8062, National Institute of Standards and Technology, ...

    2017

  31. [39]

    What should be done with equity theory? New approaches to the study of fairness in social relationships

    Gerald S Leventhal. What should be done with equity theory? New approaches to the study of fairness in social relationships. In Social exchange: Advances in theory and research, pages 27–55. Springer, 1980

    1980

  32. [40]

    Gilliland

    Stephen W. Gilliland. The Perceived Fairness of Selection Systems: An Organizational Justice Perspective. The Academy of Management Review, 18(4):694–734, 1993. Publisher: Academy of Management

    1993

  33. [41]

    Teodorescu, Yazeed Awwad, and Gerald C

    Lily Morse, Mike Horia M. Teodorescu, Yazeed Awwad, and Gerald C. Kane. Do the Ends Justify the Means? Variation in the Distributive and Procedural Fairness of Machine Learning Algorithms.Journal of Business Ethics, 181(4):1083–1095, December 2022

    2022

  34. [42]

    Redmiles, Krishna P

    Nina Grgic-Hlaca, Elissa M. Redmiles, Krishna P. Gummadi, and Adrian Weller. Human Perceptions of Fairness in Algorithmic Decision Making: A Case Study of Criminal Risk Prediction. In Proceedings of the 2018 World Wide Web Conference on World Wide Web - WWW ’18, pages 903–912,...

    2018

  35. [43]

    Big data and due process: Toward a framework to redress predictive privacy harms

    Kate Crawford and Jason Schultz. Big data and due process: Toward a framework to redress predictive privacy harms. BCL Rev., 55:93, 2014. Publisher: HeinOnline

    2014

  36. [44]

    Predicting academic performance of students from VLE big data using deep learning models

    Hajra Waheed, Saeed-Ul Hassan, Naif Radi Aljohani, Julie Hardman, Salem Alelyani, and Raheel Nawaz. Predicting academic performance of students from VLE big data using deep learning models. Computers in Human Behavior, 104:106189, 2020

    2020

  37. [45]

    A Systematic Review About Prediction of Academic Behavior Through Data Mining Techniques.Journal of Computational and Theoretical Nanoscience, 17(11):5162– 5166, November 2020

    Puninder Kaur, Amandeep Kaur, and Rajwinder Kaur. A Systematic Review About Prediction of Academic Behavior Through Data Mining Techniques.Journal of Computational and Theoretical Nanoscience, 17(11):5162– 5166, November 2020

    2020

  38. [46]

    E-cheating Prevention Measures: Detection of Cheating at Online Examinations Using Deep Learning Approach – A Case Study, January 2021

    Leslie Ching Ow Tiong and HeeJeong Jasmine Lee. E-cheating Prevention Measures: Detection of Cheating at Online Examinations Using Deep Learning Approach – A Case Study, January 2021. arXiv:2101.09841 [cs]

    2021 arXiv

  39. [47]

    Predictive Analytics in Hiring: The Key to Forecasting Workforce Needs - Blog, April 2024

    Neha Gaba. Predictive Analytics in Hiring: The Key to Forecasting Workforce Needs - Blog, April 2024

    2024

  40. [48]

    Prediction of Employee Turnover in Organizations using Machine Learning Algorithms

    Rohit Punnoose and Pankaj Ajit. Prediction of Employee Turnover in Organizations using Machine Learning Algorithms. In International Journal of Advanced Research in Artificial Intelligence, volume 5, 2016. ISSN: 21654069, 21654050 Issue: 9 Journal Abbreviation: ijarai

    2016

  41. [49]

    Should I Trust the Artificial Intelligence to Recruit? Recruiters’ Per- ceptions and Behavior When Faced With Algorithm-Based Recommendation Systems During Resume Screening

    Alain Lacroux and Christelle Martin-Lacroux. Should I Trust the Artificial Intelligence to Recruit? Recruiters’ Per- ceptions and Behavior When Faced With Algorithm-Based Recommendation Systems During Resume Screening. Frontiers in Psychology, 13, July 2022. Publisher: Frontiers

    2022

  42. [50]

    Nazanin Cassidy, Kat. U.s. job-seekers’ organizational justice perceptions of emotion ai-enabled interviews. Proceedings of the ACM on Human-Computer Interaction, 2024

    2024

  43. [51]

    Resume format, linkedin urls and other unexpected influences on ai personality prediction in hiring: Results of an audit

    Kelsey Alene, Hilke Lauren, Paul Mona, and Julia. Resume format, linkedin urls and other unexpected influences on ai personality prediction in hiring: Results of an audit. Proceedings of the 2022 AAAI/ACM Conference on AI, Ethics, and Society, 2022

    2022

  44. [52]

    Understanding Utility and Privacy of Demographic Data in Education Technology by Causal Analysis and Adversarial-Censoring

    Rakibul Hasan and Mario Fritz. Understanding Utility and Privacy of Demographic Data in Education Technology by Causal Analysis and Adversarial-Censoring. Proceedings on Privacy Enhancing Technologies, 2022(2):245– 262, April 2022

    2022

  45. [53]

    Overlearning reveals sensitive attributes.arXiv preprint arXiv:1905.11742, 2019

    Congzheng Song and Vitaly Shmatikov. Overlearning reveals sensitive attributes.arXiv preprint arXiv:1905.11742, 2019

    1905 arXiv

  46. [54]

    Self-Destructing Models: Increasing the Costs of Harmful Dual Uses of Foundation Models

    Peter Henderson, Eric Mitchell, Christopher Manning, Dan Jurafsky, and Chelsea Finn. Self-Destructing Models: Increasing the Costs of Harmful Dual Uses of Foundation Models. In Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society, AIES ’23, pages 287–296, New...

    2023

  47. [55]

    Gender, Race, and Intersectional Bias in Resume Screening via Language Model Retrieval

    Kyra Wilson and Aylin Caliskan. Gender, Race, and Intersectional Bias in Resume Screening via Language Model Retrieval. Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, 7(1):1578–1590, October 2024. Number: 1

    2024

  48. [56]

    Dissecting racial bias in an algorithm used to manage the health of populations

    Ziad Obermeyer, Brian Powers, Christine V ogeli, and Sendhil Mullainathan. Dissecting racial bias in an algorithm used to manage the health of populations. Science, 366(6464):447–453, October 2019

    2019

  49. [57]

    Toward inclusive justice: Applying the Diverse V oices design method to improve the Washington State Access to Justice Technology Principles

    Lassana Magassa and Batya Friedman. Toward inclusive justice: Applying the Diverse V oices design method to improve the Washington State Access to Justice Technology Principles. ACM J. Responsib. Comput., 1(3):18:1– 18:30, July 2024

    2024

  50. [58]

    EDM and privacy: Ethics and legalities of data collection, usage, and storage

    Mark Klose, Vasvi Desai, Yang Song, and Edward Gehringer. EDM and privacy: Ethics and legalities of data collection, usage, and storage. International Educational Data Mining Society, 2020. Publisher: ERIC

    2020

  51. [59]

    Who’s learning? Using demographics in EDM research

    Luc Paquette, Jaclyn Ocumpaugh, Ziyue Li, Alexandra Andres, and Ryan Baker. Who’s learning? Using demographics in EDM research. Journal of Educational Data Mining, 12(3):1–30, 2020. Publisher: ERIC

    2020

  52. [60]

    The development of markers for the Big-Five factor structure

    Lewis R Goldberg. The development of markers for the Big-Five factor structure. Psychological assessment, 4(1):26, 1992. Publisher: American Psychological Association

    1992

  53. [61]

    The Relationships Between Personality Traits and Students’ Academic Achievement

    Soraya Hakimi, Elaheh Hejazi, and Masoud Gholamali Lavasani. The Relationships Between Personality Traits and Students’ Academic Achievement. Procedia - Social and Behavioral Sciences, 29:836–845, January 2011

    2011

  54. [62]

    Pre-Employment Personality Tests, Algorithmic Bias, and the Americans with Disabilities Act

    Kelly Cahill Timmons. Pre-Employment Personality Tests, Algorithmic Bias, and the Americans with Disabilities Act. Penn State Law Review, 125:389, 2020

    2020

  55. [63]

    An Innovative Approach of Personality Recognition for E-Recruitment

    Priyanka Kamble and Umesh Kulkarni. An Innovative Approach of Personality Recognition for E-Recruitment. In 2022 5th International Conference on Advances in Science and Technology (ICAST), pages 324–328, December 2022. 20 A PREPRINT - S EPTEMBER 18, 2025

    2022

  56. [64]

    Van Mil, Ayushi Rastogi, and Andy Zaidman

    Frenk C.J. Van Mil, Ayushi Rastogi, and Andy Zaidman. Promises and Perils of Inferring Personality on GitHub. In Proceedings of the 15th ACM / IEEE International Symposium on Empirical Software Engineering and Measurement (ESEM), pages 1–11, Bari Italy, October 2021. ACM

    2021

  57. [65]

    Multimodal analysis of personality traits on videos of self-presentation and induced behavior

    Dersu Giritlio˘glu, Burak Mandira, Selim Firat Yilmaz, Can Ufuk Ertenli, Berhan Faruk Akgür, Merve Kınıklıo˘glu, Aslı Gül Kurt, Emre Mutlu, ¸ Seref Can Gürel, and Hamdi Dibeklio˘glu. Multimodal analysis of personality traits on videos of self-presentation and induced behavior....

    2021

  58. [66]

    Emotion AI at Work: Implications for Workplace Surveillance, Emotional Labor, and Emotional Privacy

    Kat Roemmich, Florian Schaub, and Nazanin Andalibi. Emotion AI at Work: Implications for Workplace Surveillance, Emotional Labor, and Emotional Privacy. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems, pages 1–20, Hamburg Germany, April 2023. ACM

    2023

  59. [67]

    Values in Emotion Artificial Intelligence Hiring Services: Technosolutions to Organizational Problems

    Kat Roemmich, Tillie Rosenberg, Serena Fan, and Nazanin Andalibi. Values in Emotion Artificial Intelligence Hiring Services: Technosolutions to Organizational Problems. Proceedings of the ACM on Human-Computer Interaction, 7(CSCW1):1–28, April 2023

    2023

  60. [68]

    Recruitment and selection process using artificial intelligence: How do candidates react? Administrative Sciences, 14(7):155, 2024

    Nuno Ligeiro, Ivo Dias, and Ana Moreira. Recruitment and selection process using artificial intelligence: How do candidates react? Administrative Sciences, 14(7):155, 2024. Accessed: 2025-04-10

    2024

  61. [69]

    Creativity and artificial intelligence—a student perspective

    Rebecca Marrone, Victoria Taddeo, and Gillian Hill. Creativity and artificial intelligence—a student perspective. Journal of Intelligence, 10(3):65, 2022. Accessed: 2025-04-10

    2022

  62. [70]

    Nugent, Paul Jackson, Susan Scott-Parker, James Partridge, Rebecca Raper, Alex Shepherd, Chara Bakalis, Arijit Mitra, Jintao Long, Kevin Maynard, and Nigel Crook

    Selin E. Nugent, Paul Jackson, Susan Scott-Parker, James Partridge, Rebecca Raper, Alex Shepherd, Chara Bakalis, Arijit Mitra, Jintao Long, Kevin Maynard, and Nigel Crook. Recruitment ai has a disability problem: Questions employers should be asking to ensure fairness in recru...

    2020

  63. [71]

    Tackling algorithmic disability discrimi- nation in the hiring process: An ethical, legal and technical analysis

    Maarten Buyl, Christina Cociancig, Cristina Frattone, and Nele Roekens. Tackling algorithmic disability discrimi- nation in the hiring process: An ethical, legal and technical analysis. In Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency (FA...

    2022

  64. [72]

    The Boundaries of Privacy Harm

    Ryan Calo. The Boundaries of Privacy Harm. Indiana Law Journal, 86:1131, 2011

    2011

  65. [73]

    Sayre, Ushnish Sengupta, Arthit Suriyawongkul, Ruby Thelot, Sofia Vei, and Laura Waltersdorfer

    Gavin Abercrombie, Djalel Benbouzid, Paolo Giudici, Delaram Golpayegani, Julio Hernandez, Pierre Noro, Harshvardhan Pandit, Eva Paraschou, Charlie Pownall, Jyoti Prajapati, Mark A. Sayre, Ushnish Sengupta, Arthit Suriyawongkul, Ruby Thelot, Sofia Vei, and Laura Waltersdorfer. ...

    2024 arXiv

  66. [74]

    https://gdpr-info.eu/, 2024

    General data protection regulation. https://gdpr-info.eu/, 2024

    2024

  67. [75]

    Solon Barocas and Andrew D. Selbst. Big data’s disparate impact. California Law Review, 104(3):671–732, 2016

    2016

  68. [76]

    A right to reasonable inferences: Re-thinking data protection law in the age of big data

    Sandra Wachter and Brent Mittelstadt. A right to reasonable inferences: Re-thinking data protection law in the age of big data. Columbia Business Law Review, 2019(2):494–620, 2019

    2019

  69. [77]

    Ahmad, N

    T. Ahmad, N. Ismail, and N. A. Aziz. Implications of artificial intelligence on student privacy and autonomy. Education and Information Technologies, 28:10241–10260, 2023

    2023

  70. [78]

    Technology, autonomy, and manipulation

    Daniel Susser, Beate Roessler, and Helen Nissenbaum. Technology, autonomy, and manipulation. Internet Policy Review, 8(2), 2019

    2019

  71. [79]

    Bryson, and Arvind Narayanan

    Aylin Caliskan, Joanna J. Bryson, and Arvind Narayanan. Semantics derived automatically from language corpora contain human-like biases. Science, 356(6334):183–186, 2017

    2017

  72. [80]

    Akgun and C

    S. Akgun and C. Greenhow. Artificial intelligence in education: Addressing ethical challenges in k-12 settings. British Journal of Educational Technology, 52(4):1541–1558, 2021

    2021

  73. [81]

    Predatory predictions and the ethics of predictive analytics

    Kirsten Martin. Predatory predictions and the ethics of predictive analytics. Journal of the Association for Information Science and Technology , 74(5):531–545, 2023. _eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/asi.24743

    2023 doi

  74. [82]

    Learning from ofqual’s 2020 grading algorithm failure

    Janet Hanson. Learning from ofqual’s 2020 grading algorithm failure. British Educational Research Journal, 47(3):729–737, 2021

    2020

  75. [83]

    Garfield Tenzer

    Lisa Y . Garfield Tenzer. Defamation in the age of artificial intelligence.SSRN Electronic Journal, 2023. Available at SSRN: https://ssrn.com/abstract=4567891

    2023

  76. [84]

    The malicious use of artificial intelligence: Forecasting, prevention, and mitigation

    Miles Brundage, Shahar Avin, Jack Clark, Helen Toner, et al. The malicious use of artificial intelligence: Forecasting, prevention, and mitigation. arXiv preprint arXiv:1802.07228, 2018

    2018 arXiv

  77. [85]

    Worries about artificial intelligence, surveillance at work may be connected to poor mental health, 2023

    American Psychological Association. Worries about artificial intelligence, surveillance at work may be connected to poor mental health, 2023. Press Release, Sep. 7, 2023. 21 A PREPRINT - S EPTEMBER 18, 2025

    2023

  78. [86]

    Prolific · Quickly find research participants you can trust

  79. [87]

    Making sense of Cronbach’s alpha

    Mohsen Tavakol and Reg Dennick. Making sense of Cronbach’s alpha. International Journal of Medical Education, 2:53–55, June 2011

    2011

  80. [88]

    Discovering statistics using R

    Andy Field, Zoe Field, and Jeremy Miles. Discovering statistics using R. 2012. Publisher: sage

    2012

  81. [89]

    Privacy preferences differ by gender and age, but not by income.The Technology Policy Institute, 2023

    Jeffrey Prince and Scott Wallsten. Privacy preferences differ by gender and age, but not by income.The Technology Policy Institute, 2023. Accessed: 2025-04-09

    2023

  82. [90]

    Gender differences in beliefs about algorithmic fairness

    Emma Pierson. Gender differences in beliefs about algorithmic fairness. arXiv preprint arXiv:1712.09124, 2017. Accessed: 2025-04-10

    2017 arXiv

  83. [91]

    Ekaterina Novozhilova, Kate Mays, Sejin Paik, and James E. Katz. More capable, less benevolent: Trust perceptions of ai systems across societal contexts. Machine Learning and Knowledge Extraction, 6(1):342–366, 2024

    2024

  84. [92]

    White fragility: Why Understanding racism can be so hard for white people (adapted for young adults)

    Robin DiAngelo. White fragility: Why Understanding racism can be so hard for white people (adapted for young adults). Beacon press, 2022

    2022

  85. [93]

    Privacy, accuracy, and model fairness trade-offs in federated learning

    Xiuting Gu, Zhu Tianqing, Jie Li, Tao Zhang, Wei Ren, and Kim-Kwang Raymond Choo. Privacy, accuracy, and model fairness trade-offs in federated learning. Computers & Security, 122:102907, November 2022

    2022

  86. [94]

    On the Privacy Risks of Algorithmic Fairness

    Hongyan Chang and Reza Shokri. On the Privacy Risks of Algorithmic Fairness. In 2021 IEEE European Symposium on Security and Privacy (EuroS&P), pages 292–303, September 2021

    2021

  87. [95]

    Massey and Annie I

    Aaron K. Massey and Annie I. Antón. A Requirements-based Comparison of Privacy Taxonomies. In 2008 Requirements Engineering and Law, pages 1–5, September 2008

    2008

  88. [96]

    Privacy Impact Tree Analysis (PITA): A Tree-based Privacy Threat Modeling Approach

    Dimitri Van Landuyt. Privacy Impact Tree Analysis (PITA): A Tree-based Privacy Threat Modeling Approach. IEEE Transactions on Software Engineering, pages 1–23, 2025

    2025

  89. [97]

    Trust, Because You Can’t Verify: Privacy and Security Hurdles in Education Technology Acquisition Practices

    Easton Kelso, Ananta Soneji, Sazzadur Rahaman, Yan Shoshitaishvili, and Rakibul Hasan. Trust, Because You Can’t Verify: Privacy and Security Hurdles in Education Technology Acquisition Practices. In Proceedings of the 2024 on ACM SIGSAC Conference on Computer and Communication...

    2024

  90. [98]

    Is it a Concern or a Preference? An Investigation into the Ability of Privacy Scales to Capture and Distinguish Granular Privacy Constructs

    Jessica Colnago, Lorrie Faith Cranor, Alessandro Acquisti, and Kate Hazel Jain. Is it a Concern or a Preference? An Investigation into the Ability of Privacy Scales to Capture and Distinguish Granular Privacy Constructs. In Eighteenth symposium on usable privacy and security (...

    2022

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.