Key papers in Safety Research
Pofolia’s corpus holds 83 papers from the Safety Research subfield (2012–2025). The list below starts with the most cited.
Most cited
Ranked by citation count. Because citations accumulate over time, this list naturally leans towards work published a few years ago; for where the field is now, see “recently added”.
Evaluating Online Labor Markets for Experimental Research: Amazon.com's Mechanical Turk
Political Analysis · 2012 · Q1 · SJR 4.00 · FWCI 298.95 · 4,087 citations · Open access
Amazon's Mechanical Turk (MTurk) offers a low-cost method for recruiting participants for political science experiments, yielding samples more representative of the U.S. population than typical in-person convenience samples.
AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations
Minds and Machines · 2018 · Q1 · SJR 1.00 · FWCI 168.15 · 3,679 citations · Open access
The AI4People initiative proposes a framework for a 'Good AI Society' by outlining key opportunities and risks, synthesizing five core ethical principles, and offering 20 concrete recommendations for AI development and adoption.
Post-Capitalist Society
2012 · FWCI 166.81 · 3,592 citations
Peter Drucker argues that in post-capitalist society, the primary economic resource shifts from capital, natural resources, or labor to knowledge.
The ethics of algorithms: Mapping the debate
Big Data & Society · 2016 · Q1 · SJR 2.00 · FWCI 161.09 · 3,004 citations · Open access
This paper maps the ethical debate surrounding algorithms, which increasingly mediate human decisions and social processes.
Women Empowerment and Economic Development
Journal of Economic Literature · 2012 · Q1 · SJR 16.00 · FWCI 135.84 · 2,850 citations
Economic development alone may not be enough to drive down gender inequality, and empowering women might not automatically benefit development; their interrelationships may be too weak to be self-sustaining.
Career Development and Counseling: Putting Theory and Research to Work
2020 · FWCI 202.26 · 2,697 citations
This volume is presented as an essential resource for understanding career development, assessment, and counseling, featuring significant work in the field.
How the machine ‘thinks’: Understanding opacity in machine learning algorithms
Big Data & Society · 2016 · Q1 · SJR 2.00 · FWCI 163.79 · 2,546 citations · Open access
This article distinguishes three forms of opacity in machine learning algorithms used for classification and ranking: corporate/state secrecy, technical illiteracy, and inherent algorithmic opacity. It argues that identifying the specific type of opacity is crucial for selecting appropriate solutions to prevent harm.
Women, Minorities, and Persons with Disabilities in Science and Engineering: 2013
PsycEXTRA Dataset · 2013 · 2,505 citations
This 2013 report compiles statistical data on the representation of women, minorities, and individuals with disabilities in science and engineering education and employment.
What is AI Literacy? Competencies and Design Considerations
2020 · 2,406 citations
This paper defines AI literacy by synthesizing interdisciplinary research into a set of core competencies and proposes design considerations for AI developers and educators.
Human Trust in Artificial Intelligence: Review of Empirical Research
Academy of Management Annals · 2020 · Q1 · SJR 14.00 · FWCI 161.26 · 2,356 citations
This review synthesizes 20 years of research on human trust in AI, identifying AI's representation (robot, virtual, embedded) and machine intelligence level as key antecedents. It highlights tangibility, transparency, reliability, and immediacy for cognitive trust, and anthropomorphism for emotional trust.
Measuring the Prevalence of Questionable Research Practices With Incentives for Truth Telling
Psychological Science · 2012 · Q1 · SJR 2.00 · FWCI 132.07 · 2,336 citations
A survey of over 2,000 psychologists using anonymous reporting and truth-telling incentives revealed a surprisingly high prevalence of questionable research practices, suggesting they may be the prevailing norm.
A systematic review of the psychological and social benefits of participation in sport for children and adolescents: informing development of a conceptual model of health through sport
International Journal of Behavioral Nutrition and Physical Activity · 2013 · Q1 · SJR 2.00 · FWCI 24.60 · 2,297 citations · Open access
A systematic review of 30 studies found that sport participation, particularly team sports, offers psychological and social health benefits for children and adolescents beyond general physical activity, including improved self-esteem and social interaction.
Data Collection in a Flat World: The Strengths and Weaknesses of Mechanical Turk Samples
Journal of Behavioral Decision Making · 2012 · Q1 · FWCI 165.69 · 2,294 citations
Mechanical Turk (MTurk) offers quick, inexpensive access to research participants, but exhibits key differences from traditional samples, including lower attention and a tendency to search for answers online.
Promoting Positive Youth Development Through School-Based Social and Emotional Learning Interventions: A Meta-Analysis of Follow-Up Effects
Child Development · 2017 · Q1 · SJR 1.00 · FWCI 32.90 · 2,202 citations
A meta-analysis of 82 school-based social and emotional learning (SEL) interventions shows that SEL significantly enhances positive youth development, with benefits persisting for up to 18 years post-intervention.
The age of surveillance capitalism: the fight for a human future at the new frontier of power
Journal of Cyber Policy · 2019 · 2,114 citations
This article introduces the concept of 'surveillance capitalism' as a new frontier of power, framing it as a fight for humanity's future.
European Union Regulations on Algorithmic Decision Making and a “Right to Explanation”
AI Magazine · 2017 · Q2 · FWCI 257.19 · 2,099 citations
The EU's General Data Protection Regulation (GDPR), effective 2018, will restrict automated individual decision-making by algorithms that significantly affect users.
Equality of Opportunity in Supervised Learning
arXiv (Cornell University) · 2016 · 1,926 citations · Open access
Researchers propose a new criterion for measuring and removing discrimination against sensitive attributes in supervised machine learning models, shifting the cost of poor classification to the decision-maker.
Certifying and Removing Disparate Impact
2015 · 1,799 citations
This paper defines disparate impact in algorithms as a legal concept where a seemingly neutral selection process yields widely different outcomes for different protected groups.
Model Cards for Model Reporting
2019 · 1,774 citations
This paper introduces 'model cards,' a standardized documentation framework for machine learning models to enhance transparency and guide appropriate usage.
South African Journal of Education
South African Journal of Education · 2025 · Q3 · 1,717 citations
This paper investigates whether mathematics learning facilitators in South Africa are implementing metacognitive strategies in their teaching.
Recently added
South African Journal of Education
South African Journal of Education · 2025 · Q3 · 1,717 citations
This paper investigates whether mathematics learning facilitators in South Africa are implementing metacognitive strategies in their teaching.
Behavioral Game Theory
Cambridge University Press eBooks · 2025 · 1,345 citations
Behavioral economics, enriched by psychology, sociology, evolutionary biology, and neuroscience, offers a more empirically supported foundation for economic principles and necessitates a fundamental transformation of economics.
Why Johnny Can’t Prompt: How Non-AI Experts Try (and Fail) to Design LLM Prompts
2023 · 830 citations · Open access
Non-AI experts struggle to design effective prompts for large language models (LLMs), exploring prompt designs opportunistically rather than systematically and encountering barriers like overgeneralization and expectations from human-to-human instruction.
Generative AI and the future of education: Ragnarök or reformation? A paradoxical perspective from management educators
The International Journal of Management Education · 2023 · Q1 · SJR 1.00 · FWCI 226.14 · 1,032 citations
Generative AI presents a paradox for education: it is both a 'friend' and 'foe,' 'capable' yet 'dependent,' 'accessible' but 'restrictive,' and gains popularity when 'banned.'
Collaborating With ChatGPT: Considering the Implications of Generative Artificial Intelligence for Journalism and Media Education
Journalism & Mass Communication Educator · 2023 · Q1 · SJR 1.00 · FWCI 202.89 · 928 citations · Open access
ChatGPT can generate text responses based on user prompts, demonstrating both capacity and limitations for journalism and media education.
Add this field to your daily feed
Pick your interests and new work in your area arrives every day, summarised. Full summaries live in the app.
Open the app