REVIEW 4 major objections 5 minor 88 references
Social Scientists on the Role of AI in Research
T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read When an identical survey asks about 'machine learning' instead of 'artificial intelligence,' social scientists report markedly higher acceptance, higher perceived usefulness, and less ethical concern.
desk verdict A genuinely novel randomized label comparison that deserves review, but the reporting gaps make the central causal claim provisional. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The machinery is a randomized between-subjects label split embedded in an otherwise conventional survey: every question, closed and open, was held constant and only the phrase 'artificial intelligence' versus 'machine learning' was swapped, after which 15 semi-structured interviews probed the reasons behind the survey differences. The label split is what transforms a descriptive attitude survey into a test of whether the terms themselves carry different social meaning; the paired interviews supply the mechanism, showing that researchers map 'AI' onto generative tools and 'ML' onto prediction and classification.
What would settle it
Check whether the two arms are actually comparable: a balance table across all observed covariates, a randomization log, or a nonresponse analysis showing who completed each version would settle it. If the acceptance gap (3.49 vs 2.98) vanished once respondents were given definitions of both terms, the effect would be terminological confusion rather than a settled difference in trust; a within-subject version of the same questions would reveal whether the gap survives when the same person rates both labels.
Extended reading notes
Core claim
The paper's central discovery is a label effect: the same closed-ended questions, with only the target term randomized between 'Artificial Intelligence' and 'Machine Learning,' produce statistically significant differences on every outcome measured. ML respondents reported higher acceptance (3.49 vs 2.98 mean, p = 0.0001), higher perceived usefulness (3.37 vs 2.95, p = 0.00014), and lower familiarity (2.72 vs 3.19, p = 0.00012) than AI respondents, alongside differences in usage frequency. The authors' interpretation is that 'AI' has come to mean generative, opaque, black-box systems in the social science community, while 'ML' still carries the connotation of rigorous, transparent, statistically grounded methods; the lower familiarity reported for ML is consistent with ML's narrower, more technical association. Open-ended ethical concerns also appeared more frequently and more intensely in the AI-survey arm. The discovery's significance is that perceived trust and ethical concern are attached to the label as much as to the underlying technology.
Load-bearing premise
The claim that the label itself changes trust rests on the assumption that the two groups of respondents are otherwise alike: the split between 'AI' and 'ML' surveys was truly random, and the people who happened to complete each version were equally willing and able to answer, so the only systematic difference between the groups was the word on the questionnaire.
Editorial extensions
If this is right
- Surveys, policy documents, and grant guidance that use 'AI' and 'machine learning' interchangeably will systematically misread community sentiment: a tool presented as ML can expect roughly 16 percentage points more high acceptance than the same tool presented as AI.
- Ethical concern is concentrated on generative, opaque systems rather than on predictive modeling, so governance should target genAI-specific failure modes — automation bias, fabrication, representational harm, and environmental cost.
- Adoption and acceptance divide by career stage and gender: PhD students are the most accepting of AI, while female researchers report lower usage and familiarity, pointing to where training and capacity-building should be aimed.
- Social scientists widely endorse AI as a collaborator for literature review, coding, and annotation but insist that validation, theory-building, and ethical judgment remain human tasks, a stance that argues for human-in-the-loop design rather than full automation.
- The absence of standardization in documenting AI use is a barrier to trust and reproducibility; the paper's proposed documentation protocols (model transparency, usage disclosure, evaluation criteria, reproducibility artifacts) are the concrete response.
Reading between the lines
- If the label effect is real, it should generalize beyond this sample: a cheap, decisive replication would re-administer the same two-arm survey to a different population — grant reviewers, journal editors, IRB members — whose labeling choices shape what research gets funded and approved.
- The familiarity paradox (AI respondents report higher familiarity yet lower trust) suggests the gap tracks public discourse about generative AI rather than technical understanding; a within-subject follow-up giving each respondent a definition of both terms before asking the questions would show whether the effect is about the words themselves or about what the words evoke.
- The causal reading of the comparison rests on the two arms being genuinely comparable; a fuller randomization and nonresponse report, or a covariate balance table covering more than demographics, would harden the result. Without it, part of the 51.77% versus 35.66% gap could be differential response propensity rather than the label.
- For the field, the practical upshot is that 'AI' and 'ML' should be treated as a measured variable in every future attitude survey, not as a fixed instrument choice; meta-analyses of researcher attitudes should code for the term used.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper reports a mixed-methods study of social scientists' familiarity with, use of, acceptance of, and ethical concerns about AI versus machine learning (ML). The authors surveyed 284 U.S.-based social science researchers, randomly splitting the sample so that half answered questions about "AI" and half about "Machine Learning," and conducted 15 follow-up interviews. The central finding is that respondents evaluated ML more favorably than AI on acceptance and usefulness, despite reporting lower familiarity with ML, and that interview participants associated AI with generative, black-box systems and ML with statistically grounded methods. The paper also catalogs concerns about deskilling, bias, interpretability, reproducibility, and environmental harm, and offers stakeholder recommendations. The abstract further claims that AI use among social scientists "has increased significantly" in step with generative AI, and the authors interpret the label effect as evidence that the terms themselves carry different social and technical connotations.
Significance. If the label effect is real, the study makes a valuable and falsifiable contribution: it shows that survey responses about "AI" versus "machine learning" differ even when the questions are otherwise identical, which has direct implications for how social scientists and policymakers measure attitudes toward these technologies. The randomized-label design is a genuine innovation in this literature, and the mixed-methods structure provides rich contextual evidence that the quantitative differences are not arbitrary. The paper also addresses a timely and important question about how researchers negotiate trust, transparency, and human judgment as generative AI becomes embedded in academic work. The qualitative data on the distinction between AI-as-product and ML-as-method is thought-provoking. However, the central causal claim rests on the assumption of successful random assignment and comparable response propensities across the two survey arms, and the manuscript does not yet provide the design and balance details needed to support that assumption. The cross-sectional design also cannot by itself support the temporal claim of increased use.
major comments (4)
- [Methods, Participant recruitment and compensation] The central claim that the label "AI" versus "ML" changes responses requires that assignment to survey version be randomized and that response propensity be comparable across arms, but the manuscript does not describe the randomization mechanism, whether assignment was at the individual or university level, or how the stated split of 8,000 invited researchers was implemented. Because invitations were scraped from departmental pages at 30 universities, assignment could have been clustered; if so, university-level differences in field mix or departmental culture could confound the Table 2 comparisons. The authors should report the randomization procedure, present balance tests on the full invited sample where possible, and show balance statistics (not just raw counts) for the 284 completers.
- [Abstract and RQ1 section] The statement in the abstract that AI use "has increased significantly" among social scientists is not supported by the data presented. The survey is cross-sectional and asks about current frequency of use; no longitudinal or retrospective measure of change is reported. The interview participants describe increased use, but that is self-reported impression among a small, nonrepresentative sample. The authors should either present a direct measure of change over time or temper the claim to reflect current usage patterns and perceived increases.
- [Table 2 and 'Usefulness and acceptance'] The four p-values in Table 2 are reported without stating the statistical test used, without effect sizes or confidence intervals, and without any adjustment for multiple comparisons. The text in the 'Usefulness and acceptance' subsection says the differences were statistically significant (p < .01), but Table 2 shows p = 0.046 for frequency of use, so the summary is not uniformly accurate. The authors should specify the test for each row, report effect sizes, and either correct for multiple testing or explicitly frame the results as exploratory.
- [Participant demographics and data analysis] The response rate is 3.56%, yet no nonresponse analysis is provided and the only balance assessment is a verbal statement that demographics are similar across arms. With such a low completion rate, differential nonresponse is a serious threat: researchers who feel negatively about AI may be more likely to abandon an AI-labeled survey, which would bias the comparison even if demographics are balanced. The authors should report nonresponse patterns by arm (e.g., proportion of incomplete responses per arm) and, if possible, compare characteristics of completers versus non-completers on the invited sample.
minor comments (5)
- [Fig 2 and 'RQ1: AI and ML reshaping social science research practices'] The eight regressions summarized in Figure 2 are not fully specified: the manuscript does not state whether the models are linear or ordered, which covariates were included, or whether standard errors account for clustering by university. Please add this information to the Methods or figure caption.
- [Participant recruitment and compensation] The text says 30 universities were selected randomly from the top 100 based on the 2003 US News and World Report rankings. Using rankings that are over twenty years old may affect generalizability and should be justified or updated.
- [Table 3 and 'Interview design'] The interview sample is heavily weighted toward PhD students (10 of 15 in Table 3). Since the interviews are used to "validate" the survey findings, this imbalance should be acknowledged as a limitation or the authors should discuss how it shapes the interpretations they draw from interviews.
- [Throughout] There are several typographical and formatting issues, including "inverviews" in the RQ2 section, "ways in which" in the Conclusion, and malformed author names in the References (e.g., "Obreja, Rughinis , , and Rosner 2024"). These should be cleaned up before publication.
- [Figure 1 and Figure 2] Figure 1 does not include confidence intervals or significance indicators, and Figure 2 appears to show point estimates without a description of the underlying model or error bars. Adding these details would make the visual claims easier to verify against Table 2.
Circularity Check
No circular reasoning: the paper's claims rest on independently collected survey and interview data, with self-citations used only as background and not as load-bearing evidence.
full rationale
This paper is an empirical, mixed-methods study. Its central claims—that social scientists perceive 'AI' and 'ML' differently on identical survey questions, and that these differences track associations with generative AI versus statistical methods—are supported by newly collected survey responses (n=284) and semi-structured interviews (n=15). The analysis involves descriptive statistics, regression analyses of demographic predictors, and thematic qualitative coding; there are no mathematical derivations, fitted parameters, or predictive models whose outputs reduce to their inputs. No equation is defined in terms of another, and no fitted value is renamed as a prediction. The paper does include several self-citations (e.g., Chakravorti, Koneru, and Rajtmajer 2025; Narayanan Venkit et al. 2023; Venkit et al. 2024; Ghosh et al. 2024), but these are used as contextual background or as examples of prior work on related topics, not as justification for the study's main empirical findings. The label-effect conclusion is not derived from any cited theorem or prior result; it is a direct comparison of survey responses across randomized arms. The absence of a detailed randomization mechanism and the low response rate (3.56%) are legitimate threats to causal validity, but they are concerns about internal validity or statistical inference, not circularity. No part of the paper's argument assumes what it seeks to prove, and no self-referential chain forces the conclusions. Therefore, the appropriate circularity score is 0.
Assumptions & free parameters
assumptions (3)
- domain assumption Respondent assignment to the AI and ML survey arms was random and produced comparable groups.
- domain assumption Self-reported familiarity, frequency, acceptance, and usefulness correspond to actual researcher behavior.
- domain assumption Thematic coding by two authors is reliable and complete.
Cite this review
Pith. "Pith review of Social Scientists on the Role of AI in Research." pith.science (2026). https://pith.science/paper/422T4DAQ
@misc{pith2026250611255,
author = {Pith},
title = {Pith review of: Social Scientists on the Role of AI in Research},
year = {2026},
howpublished = {\url{https://pith.science/paper/422T4DAQ}},
note = {Machine review of arXiv:2506.11255}
}
read the original abstract
The integration of artificial intelligence (AI) into social science research practices raises significant technological, methodological, and ethical issues. We present a community-centric study drawing on 284 survey responses and 15 semi-structured interviews with social scientists, describing their familiarity with, perceptions of the usefulness of, and ethical concerns about the use of AI in their field. A crucial innovation in study design is to split our survey sample in half, providing the same questions to each -- but randomizing whether participants were asked about "AI" or "Machine Learning" (ML). We find that the use of AI in research settings has increased significantly among social scientists in step with the widespread popularity of generative AI (genAI). These tools have been used for a range of tasks, from summarizing literature reviews to drafting research papers. Some respondents used these tools out of curiosity but were dissatisfied with the results, while others have now integrated them into their typical workflows. Participants, however, also reported concerns with the use of AI in research contexts. This is a departure from more traditional ML algorithms which they view as statistically grounded. Participants express greater trust in ML, citing its relative transparency compared to black-box genAI systems. Ethical concerns, particularly around automation bias, deskilling, research misconduct, complex interpretability, and representational harm, are raised in relation to genAI. To guide this transition, we offer recommendations for AI developers, researchers, educators, and policymakers focusing on explainability, transparency, ethical safeguards, sustainability, and the integration of lived experiences into AI design and evaluation processes.
Figures
Reference graph
Works this paper leans on
-
[1]
Alfrink, K.; Keller, I.; Kortuem, G.; and Doorn, N. 2023. Contestable AI by design: Towards a framework. Minds and Machines, 33(4): 613--639
2023
-
[2]
N.; Inkpen, K.; et al
Amershi, S.; Weld, D.; Vorvoreanu, M.; Fourney, A.; Nushi, B.; Collisson, P.; Suh, J.; Iqbal, S.; Bennett, P. N.; Inkpen, K.; et al. 2019. Guidelines for human-AI interaction. In Proceedings of the 2019 chi conference on human factors in computing systems, 1--13
2019
-
[3]
Anthis, J. R.; Liu, R.; Richardson, S. M.; Kozlowski, A. C.; Koch, B.; Evans, J.; Brynjolfsson, E.; and Bernstein, M. 2025. LLM Social Simulations Are a Promising Research Method. arXiv preprint arXiv:2504.02234
arXiv 2025
-
[4]
Athey, S.; and Imbens, G. W. 2017. The state of applied econometrics: Causality and policy evaluation. Journal of Economic perspectives, 31(2): 3--32
2017
-
[5]
Bail, C. 2023. Can generative artificial intelligence improve social science. SocArXiv
2023
-
[6]
Bail, C. A. 2024. Can Generative AI improve social science? Proceedings of the National Academy of Sciences, 121(21): e2314021121
2024
-
[7]
Baker, M. 2016. Reproducibility crisis. nature, 533(26): 353--66
2016
-
[8]
M.; and Friedman, B
Bender, E. M.; and Friedman, B. 2018. Data statements for natural language processing: Toward mitigating system bias and enabling better science. Transactions of the Association for Computational Linguistics, 6: 587--604
2018
Show all 88 references
-
[9]
M.; Gebru, T.; McMillan-Major, A.; and Shmitchell, S
Bender, E. M.; Gebru, T.; McMillan-Major, A.; and Shmitchell, S. 2021. On the dangers of stochastic parrots: Can language models be too big? In Proceedings of the 2021 ACM conference on fairness, accountability, and transparency, 610--623
2021
-
[10]
Binns, R. 2018. Fairness in machine learning: Lessons from political philosophy. In Conference on fairness, accountability and transparency, 149--159. PMLR
2018
-
[11]
Blandford, A.; Furniss, D.; and Makri, S. 2016. Qualitative HCI research: Going behind the scenes. Morgan & Claypool Publishers
2016
-
[12]
L.; Liao, Q
Blodgett, S. L.; Liao, Q. V.; Olteanu, A.; Mihalcea, R.; Muller, M.; Scheuerman, M. K.; Tan, C.; and Yang, Q. 2022. Responsible language technologies: Foreseeing and mitigating harms. In CHI Conference on Human Factors in Computing Systems Extended Abstracts, 1--3
2022
-
[13]
Y.; Saligrama, V.; and Kalai, A
Bolukbasi, T.; Chang, K.-W.; Zou, J. Y.; Saligrama, V.; and Kalai, A. T. 2016. Man is to computer programmer as woman is to homemaker? debiasing word embeddings. Advances in neural information processing systems, 29
2016
-
[14]
Boyd, D. 2021. Data, Algorithms, Fairness, Accountability. TRT Akademi, 6(11): 202--209
2021
-
[15]
Castelvecchi, D. 2016. Can we open the black box of AI? Nature News, 538(7623): 20
2016
-
[16]
a rvel \
Celik, I.; Dindar, M.; Muukkonen, H.; and J \"a rvel \"a , S. 2022. The promises and challenges of artificial intelligence for teachers: A systematic review of research. TechTrends, 66(4): 616--630
2022
-
[17]
L.; and Rajtmajer, S
Chakravorti, T.; Fraleigh, R.; Fritton, T.; McLaughlin, M.; Singh, V.; Griffin, C.; Kwasnica, A.; Pennock, D.; Giles, C. L.; and Rajtmajer, S. 2023. A prototype hybrid prediction market for estimating replicability of published work. In HHAI 2023: Augmenting Human Intellect, 3...
2023
-
[18]
Chakravorti, T.; Koneru, S.; and Rajtmajer, S. 2025. Reproducibility and replicability in research: What 452 professors think in Universities across the USA and India. PloS one, 20(3): e0319334
2025
-
[19]
Chernozhukov, V.; Chetverikov, D.; Demirer, M.; Duflo, E.; Hansen, C.; Newey, W.; and Robins, J. 2018. Double/debiased machine learning for treatment and structural parameters
2018
-
[20]
Chubb, J.; Cowling, P.; and Reed, D. 2022. Speeding up to keep up: exploring the use of AI in the research process. AI & society, 37(4): 1439--1457
2022
-
[21]
Dandamudi, S. R. P.; Sajja, J.; and Khanna, A. 2025. Advancing Cybersecurity and Data Networking Through Machine Learning-Driven Prediction Models. International Journal of Innovative Research in Computer Science and Technology, 13(1): 26--33
2025
-
[22]
Colonial Impulse
Das, D.; Guha, S.; Brubaker, J.; and Semaan, B. 2024. The" Colonial Impulse" of Natural Language Processing: An Audit of Bengali Sentiment Analysis Tools and Their Identity-based Biases. arXiv preprint arXiv:2401.10535
2024 arXiv
-
[23]
Deranty, J.-P.; and Corbin, T. 2024. Artificial intelligence and work: a critical review of recent research from the social sciences. Ai & Society, 39(2): 675--691
2024
-
[24]
Dev, S.; Sheng, E.; Zhao, J.; Amstutz, A.; Sun, J.; Hou, Y.; Sanseverino, M.; Kim, J.; Nishi, A.; Peng, N.; et al. 2022. On Measures of Biases and Harms in NLP. In Findings of the Association for Computational Linguistics: AACL-IJCNLP 2022, 246--267
2022
-
[25]
J.; Simmons, J
Dietvorst, B. J.; Simmons, J. P.; and Massey, C. 2015. Algorithm aversion: people erroneously avoid algorithms after seeing them err. Journal of experimental psychology: General, 144(1): 114
2015
-
[26]
As Uploaders, We Have the Responsibility
Ding, X.; Kou, Y.; Xu, Y.; and Zhang, P. 2022. “As Uploaders, We Have the Responsibility”: Individualized Professionalization of Bilibili Uploaders. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems, 1--14
2022
-
[27]
Dolata, M.; Feuerriegel, S.; and Schwabe, G. 2022. A sociotechnical view of algorithmic fairness. Information Systems Journal, 32(4): 754--818
2022
-
[28]
Dominguez-Olmedo, R.; Hardt, M.; and Mendler-D \"u nner, C. 2024. Questioning the survey responses of large language models. Advances in Neural Information Processing Systems, 37: 45850--45878
2024
-
[29]
Edelmann, A.; Wolff, T.; Montagne, D.; and Bail, C. A. 2020. Computational social science and sociology. Annual review of sociology, 46(1): 61--81
2020
-
[30]
V.; Muller, M.; Riedl, M
Ehsan, U.; Liao, Q. V.; Muller, M.; Riedl, M. O.; and Weisz, J. D. 2021. Expanding explainability: Towards social transparency in ai systems. In Proceedings of the 2021 CHI conference on human factors in computing systems, 1--19
2021
-
[31]
Ehsan, U.; and Riedl, M. O. 2020. Human-centered explainable ai: Towards a reflective sociotechnical approach. In HCI International 2020-Late Breaking Papers: Multimodality and Intelligence: 22nd HCI International Conference, HCII 2020, Copenhagen, Denmark, July 19--24, 2020, ...
2020
-
[32]
Einav, L.; and Levin, J. 2014. Economics in the age of big data. Science, 346(6210): 1243089
2014
-
[33]
Y.; and Mostafavi, A
Fan, C.; Xu, J.; Natarajan, B. Y.; and Mostafavi, A. 2023. Interpretable machine learning learns complex interactions of urban features to understand socio-economic inequality. Computer-Aided Civil and Infrastructure Engineering, 38(14): 2013--2029
2023
-
[34]
N.; and Ghosh, S
Gautam, S.; Venkit, P. N.; and Ghosh, S. 2024. From melting pots to misrepresentations: Exploring harms in generative ai. arXiv preprint arXiv:2403.10776
2024 arXiv
-
[35]
N.; Gautam, S.; Wilson, S.; and Caliskan, A
Ghosh, S.; Venkit, P. N.; Gautam, S.; Wilson, S.; and Caliskan, A. 2024. Do Generative AI Models Output Harm while Representing Non-Western Cultures: Evidence from A Community-Centered Approach. In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society, volume 7, 476--489
2024
-
[36]
Gillespie, T. 2014. The relevance of algorithms. Media technologies: Essays on communication, materiality, and society, 167(2014): 167
2014
-
[37]
Green, B.; and Chen, Y. 2019. Disparate interactions: An algorithm-in-the-loop analysis of fairness in risk assessments. In Proceedings of the conference on fairness, accountability, and transparency, 90--99
2019
-
[38]
Grimmer, J. 2015. We are all social scientists now: How big data, machine learning, and causal inference work together. PS: Political Science & Politics, 48(1): 80--83
2015
-
[39]
E.; and Stewart, B
Grimmer, J.; Roberts, M. E.; and Stewart, B. M. 2021. Machine learning for social science: An agnostic approach. Annual Review of Political Science, 24(1): 395--419
2021
-
[40]
C.; Christakis, N
Grossmann, I.; Feinberg, M.; Parker, D. C.; Christakis, N. A.; Tetlock, P. E.; and Cunningham, W. A. 2023. AI and the transformation of social science research. Science, 380(6650): 1108--1109
2023
-
[41]
Hacker, P. 2021. A legal framework for AI training data—from first principles to the Artificial Intelligence Act. Law, innovation and technology, 13(2): 257--301
2021
-
[42]
M.; Watts, D
Hofman, J. M.; Watts, D. J.; Athey, S.; Garip, F.; Griffiths, T. L.; Kleinberg, J.; Margetts, H.; Mullainathan, S.; Salganik, M. J.; Vazire, S.; et al. 2021. Integrating explanation and prediction in computational social science. Nature, 595(7866): 181--188
2021
-
[43]
You Don't Have To Know My Past
Huang, X.; Vitak, J.; and Tausczik, Y. 2020. " You Don't Have To Know My Past": How WeChat Moments Users Manage Their Evolving Self-Presentation. In Proceedings of the 2020 CHI Conference on Human Factors in Computing systems, 1--13
2020
-
[44]
Iorga, D.; Bocse, B.; and Olar, D. 2023. SOCIOLOGICAL ARTIFICIAL INTELLIGENCE RESEARCH (SAIR). A REVIEW AND CONCEPTUAL FRAMEWORK. Revista Romana de Sociologie
2023
-
[45]
V.; Creswell, J
Ivankova, N. V.; Creswell, J. W.; and Stick, S. L. 2006. Using mixed-methods sequential explanatory design: From theory to practice. Field methods, 18(1): 3--20
2006
-
[46]
J.; Jung, S.-g.; and Salminen, J
Jansen, B. J.; Jung, S.-g.; and Salminen, J. 2023. Employing large language models in survey research. Natural Language Processing Journal, 4: 100020
2023
-
[47]
Kaur, P.; Stoltzfus, J.; and Yellapu, V. 2018. Descriptive statistics. International Journal of Academic Medicine, 4(1): 60--63
2018
-
[48]
K.; Rand, K.; Bassett, J.; Galaitsi, S
Ligo, A. K.; Rand, K.; Bassett, J.; Galaitsi, S. E.; Trump, B. D.; Jayabalasingham, B.; Collins, T.; and Linkov, I. 2021. Comparing the emergence of technical and social sciences research in artificial intelligence. Frontiers in Computer Science, 3: 653235
2021
-
[49]
Mehrabi, N.; Morstatter, F.; Saxena, N.; Lerman, K.; and Galstyan, A. 2021. A survey on bias and fairness in machine learning. ACM computing surveys (CSUR), 54(6): 1--35
2021
-
[50]
Miller, T. 2019. Explanation in artificial intelligence: Insights from the social sciences. Artificial intelligence, 267: 1--38
2019
-
[51]
Narayanan Venkit, P. 2023. Towards a holistic approach: Understanding sociodemographic biases in nlp models using an interdisciplinary lens. In Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society, 1004--1005
2023
-
[52]
Narayanan Venkit, P.; Gautam, S.; Panchanadikar, R.; Huang, T.-H.; and Wilson, S. 2023. Unmasking nationality bias: A study of human perception of nationalities in ai-generated articles. In Proceedings of the 2023 AAAI/ACM Conference on AI, Ethics, and Society, 554--565
2023
-
[53]
Ntoutsi, E.; Fafalios, P.; Gadiraju, U.; Iosifidis, V.; Nejdl, W.; Vidal, M.-E.; Ruggieri, S.; Turini, F.; Papadopoulos, S.; Krasanakis, E.; et al. 2020. Bias in data-driven artificial intelligence systems—An introductory survey. Wiley Interdisciplinary Reviews: Data Mining an...
2020
-
[54]
M.; Rughiniș, R.; and Rosner, D
Obreja, D. M.; Rughiniș, R.; and Rosner, D. 2024. Mapping the conceptual structure of innovation in artificial intelligence research: A bibliometric analysis and systematic literature review. Journal of Innovation & Knowledge, 9(1): 100465
2024
-
[55]
Omran Zailuddin, M. F. N.; Nik Harun, N. A.; Abdul Rahim, H. A.; Kamaruzaman, A. F.; Berahim, M. H.; Harun, M. H.; and Ibrahim, Y. 2024. Redefining creative education: a case study analysis of AI in design courses. Journal of Research in Innovative Teaching & Learning, 17(2): 282--296
2024
-
[56]
O'Neil, C. 2017. Weapons of math destruction: How big data increases inequality and threatens democracy. Crown
2017
-
[57]
D.; Bender, E
Paullada, A.; Raji, I. D.; Bender, E. M.; Denton, E.; and Hanna, A. 2021. Data and its (dis) contents: A survey of dataset development and use in machine learning research. Patterns, 2(11)
2021
-
[58]
Reckinger, R.; and Wille, C. 2018. Situative interdisciplinarity: empirical reflections on ten years of cross-disciplinary research. Academic Journal of Interdisciplinary Studies, 7(3)
2018
-
[59]
D.; Whaling, K.; Carrier, L
Rosen, L. D.; Whaling, K.; Carrier, L. M.; Cheever, N. A.; and Rokkum, J. 2013. The media and technology usage and attitudes scale: An empirical investigation. Computers in human behavior, 29(6): 2501--2511
2013
-
[60]
N.; Ashaolu, T
Sadiku, M. N.; Ashaolu, T. J.; Ajayi-Majebi, A.; and Musa, S. M. 2021. Artificial intelligence in social media. International Journal of Scientific Advances, 2(1): 15--20
2021
-
[61]
Salganik, M. J. 2019. Bit by bit: Social research in the digital age. Princeton University Press
2019
-
[62]
Everyone wants to do the model work, not the data work
Sambasivan, N.; Kapania, S.; Highfill, H.; Akrong, D.; Paritosh, P.; and Aroyo, L. M. 2021. “Everyone wants to do the model work, not the data work”: Data Cascades in High-Stakes AI. In proceedings of the 2021 CHI Conference on Human Factors in Computing Systems, 1--15
2021
-
[63]
K.; Nekrutenko, A.; Taylor, J.; and Hovig, E
Sandve, G. K.; Nekrutenko, A.; Taylor, J.; and Hovig, E. 2013. Ten simple rules for reproducible computational research. PLoS computational biology, 9(10): e1003285
2013
-
[64]
Sandvig, C.; Hamilton, K.; Karahalios, K.; and Langbort, C. 2014. Auditing algorithms: Research methods for detecting discrimination on internet platforms. Data and discrimination: converting critical concerns into productive inquiry, 22(2014): 4349--4357
2014
-
[65]
A.; and Etzioni, O
Schwartz, R.; Dodge, J.; Smith, N. A.; and Etzioni, O. 2020. Green ai. Communications of the ACM, 63(12): 54--63
2020
-
[66]
Seaver, N. 2017. Algorithms as culture: Some tactics for the ethnography of algorithmic systems. Big data & society, 4(2): 2053951717738104
2017
-
[67]
D.; Boyd, D.; Friedler, S
Selbst, A. D.; Boyd, D.; Friedler, S. A.; Venkatasubramanian, S.; and Vertesi, J. 2019. Fairness and abstraction in sociotechnical systems. In Proceedings of the conference on fairness, accountability, and transparency, 59--68
2019
-
[68]
Shmueli, G. 2010. To explain or to predict?
2010
-
[69]
Star, S. L. 1999. The ethnography of infrastructure. American behavioral scientist, 43(3): 377--391
1999
-
[70]
Strubell, E.; Ganesh, A.; and McCallum, A. 2020. Energy and policy considerations for modern deep learning research. In Proceedings of the AAAI conference on artificial intelligence, volume 34, 13693--13696
2020
-
[71]
Suchman, L. A. 2007. Human-machine reconfigurations: Plans and situated actions. Cambridge university press
2007
-
[72]
Thakkar, D.; Ismail, A.; Kumar, P.; Hanna, A.; Sambasivan, N.; and Kumar, N. 2022. When is machine learning data good?: Valuing in public health datafication. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems, 1--16
2022
-
[73]
Van Noorden, R.; and Perkel, J. M. 2023. AI and science: what 1,600 researchers think. Nature, 621(7980): 672--675
2023
-
[74]
A.; Siddarth, D.; Seshadri, V.; Bali, K.; and Vashistha, A
Varanasi, R. A.; Siddarth, D.; Seshadri, V.; Bali, K.; and Vashistha, A. 2022. Feeling Proud, Feeling Embarrassed: Experiences of Low-income Women with Crowd Work. In Proceedings of the 2022 CHI Conference on Human Factors in Computing Systems, 1--18
2022
-
[75]
Varian, H. R. 2014. Big data: New tricks for econometrics. Journal of economic perspectives, 28(2): 3--28
2014
-
[76]
N.; Chakravorti, T.; Gupta, V.; Biggs, H.; Srinath, M.; Goswami, K.; Rajtmajer, S.; and Wilson, S
Venkit, P. N.; Chakravorti, T.; Gupta, V.; Biggs, H.; Srinath, M.; Goswami, K.; Rajtmajer, S.; and Wilson, S. 2024. " Confidently Nonsensical?'': A Critical Survey on the Perspectives and Challenges of'Hallucinations' in NLP. arXiv preprint arXiv:2404.07461
2024 arXiv
-
[77]
Von Eschenbach, W. J. 2021. Transparency and the black box problem: Why we do not trust AI. Philosophy & Technology, 34(4): 1607--1622
2021
-
[78]
H.; Stanton, N
Walker, G. H.; Stanton, N. A.; Salmon, P. M.; and Jenkins, D. P. 2008. A review of sociotechnical systems theory: a classic concept for new command and control paradigms. Theoretical issues in ergonomics science, 9(6): 479--499
2008
-
[79]
Williams, R. T. 2024. Paradigm shifts: exploring AI's influence on qualitative inquiry and analysis. Frontiers in Research Metrics and Analytics, 9: 1331589
2024
-
[80]
M.; and Rajtmajer, S
Wu, C.; Chakravorti, T.; Carroll, J. M.; and Rajtmajer, S. 2024 a . Integrating measures of replicability into scholarly search: Challenges and opportunities. In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems, 1--18
2024
-
[81]
Wu, C.; Wang, X.; Carroll, J.; and Rajtmajer, S. 2024 b . Reacting to generative AI: Insights from student and faculty discussions on Reddit. In Proceedings of the 16th ACM Web Science Conference, 103--113
2024
-
[82]
Wulff, P.; Buschh \"u ter, D.; Westphal, A.; Mientus, L.; Nowak, A.; and Borowski, A. 2022. Bridging the gap between qualitative and quantitative assessment in science education research with machine learning—A case for pretrained language models-based clustering. Journal of S...
2022
-
[83]
Xu, Y.; Liu, X.; Cao, X.; Huang, C.; Liu, E.; Qian, S.; Liu, X.; Wu, Y.; Dong, F.; Qiu, C.-W.; et al. 2021. Artificial intelligence: A powerful paradigm for scientific research. The Innovation, 2(4)
2021
-
[84]
Zhang, Y. 2023. Generative AI has lowered the barriers to computational social sciences. arXiv preprint arXiv:2311.10833
2023 arXiv
-
[85]
V.; and Bellamy, R
Zhang, Y.; Liao, Q. V.; and Bellamy, R. K. 2020. Effect of confidence and explanation on accuracy and trust calibration in AI-assisted decision making. In Proceedings of the 2020 conference on fairness, accountability, and transparency, 295--305
2020
-
[86]
Ziems, C.; Held, W.; Shaikh, O.; Chen, J.; Zhang, Z.; and Yang, D. 2024. Can large language models transform computational social science? Computational Linguistics, 50(1): 237--291
2024
-
[87]
, " * write output.state after.block = add.period write newline
ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...
-
[88]
write newline
" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.