| dc.identifier.citation | @Article{rozado2023politicalcompass, AUTHOR = {Rozado, David}, TITLE = {The Political Biases of ChatGPT}, JOURNAL = {Social Sciences}, VOLUME = {12}, YEAR = {2023}, NUMBER = {3}, ARTICLE-NUMBER = {148}, URL = {https://www.mdpi.com/2076-0760/12/3/148}, ISSN = {2076-0760}, ABSTRACT = {Recent advancements in Large Language Models (LLMs) suggest imminent commercial applications of such AI systems where they will serve as gateways to interact with technology and the accumulated body of human knowledge. The possibility of political biases embedded in these models raises concerns about their potential misusage. In this work, we report the results of administering 15 different political orientation tests (14 in English, 1 in Spanish) to a state-of-the-art Large Language Model, the popular ChatGPT from OpenAI. The results are consistent across tests; 14 of the 15 instruments diagnose ChatGPT answers to their questions as manifesting a preference for left-leaning viewpoints. When asked explicitly about its political preferences, ChatGPT often claims to hold no political opinions and to just strive to provide factual and neutral information. It is desirable that public facing artificial intelligence systems provide accurate and factual information about empirically verifiable issues, but such systems should strive for political neutrality on largely normative questions for which there is no straightforward way to empirically validate a viewpoint. Thus, ethical AI systems should present users with balanced arguments on the issue at hand and avoid claiming neutrality while displaying clear signs of political bias in their content.}, DOI = {10.3390/socsci12030148} } @inproceedings{agarwal2022graphnli, title={GraphNLI: A Graph-based Natural Language Inference Model for Polarity Prediction in Online Debates}, author={Agarwal, Vibhor and Joglekar, Sagar and Young, Anthony P and Sastry, Nishanth}, booktitle={Proceedings of the ACM Web Conference 2022}, pages={2729--2737}, year={2022} } @article{agarwal2022graph, title={A Graph-Based Context-Aware Model to Understand Online Conversations}, author={Agarwal, Vibhor and Young, Anthony P and Joglekar, Sagar and Sastry, Nishanth}, journal={arXiv preprint arXiv:2211.09207}, year={2022} } @misc{zhu2023chatgpt, title={Can ChatGPT Reproduce Human-Generated Labels? A Study of Social Computing Tasks}, author={Yiming Zhu and Peixian Zhang and Ehsan-Ul Haq and Pan Hui and Gareth Tyson}, year={2023}, eprint={2304.10145}, archivePrefix={arXiv}, primaryClass={cs.AI} } @inproceedings{si_why_2022, address = {Los Angeles CA USA}, title = {Why {So} {Toxic}?: {Measuring} and {Triggering} {Toxic} {Behavior} in {Open}-{Domain} {Chatbots}}, isbn = {978-1-4503-9450-5}, shorttitle = {Why {So} {Toxic}?}, url = {https://dl.acm.org/doi/10.1145/3548606.3560599}, doi = {10.1145/3548606.3560599}, abstract = {Chatbots are used in many applications, e.g., automated agents, smart home assistants, interactive characters in online games, etc. Therefore, it is crucial to ensure they do not behave in undesired manners, providing offensive or toxic responses to users. This is not a trivial task as state-of-the-art chatbot models are trained on large, public datasets openly collected from the Internet. This paper presents a first-of-its-kind, large-scale measurement of toxicity in chatbots. We show that publicly available chatbots are prone to providing toxic responses when fed toxic queries. Even more worryingly, some non-toxic queries can trigger toxic responses too. We then set out to design and experiment with an attack, ToxicBuddy, which relies on fine-tuning GPT-2 to generate non-toxic queries that make chatbots respond in a toxic manner. Our extensive experimental evaluation demonstrates that our attack is effective against public chatbot models and outperforms manually-crafted malicious queries proposed by previous work. We also evaluate three defense mechanisms against ToxicBuddy, showing that they either reduce the attack performance at the cost of affecting the chatbot’s utility or are only effective at mitigating a portion of the attack. This highlights the need for more research from the computer security and online safety communities to ensure that chatbot models do not hurt their users. Overall, we are confident that ToxicBuddy can be used as an auditing tool and that our work will pave the way toward designing more effective defenses for chatbot safety.}, language = {en}, urldate = {2023-03-17}, booktitle = {Proceedings of the 2022 {ACM} {SIGSAC} {Conference} on {Computer} and {Communications} {Security}}, publisher = {ACM}, author = {Si, Wai Man and Backes, Michael and Blackburn, Jeremy and De Cristofaro, Emiliano and Stringhini, Gianluca and Zannettou, Savvas and Zhang, Yang}, month = nov, year = {2022}, pages = {2659--2673}, } @inproceedings{abid_persistent_2021, address = {Virtual Event USA}, title = {Persistent {Anti}-{Muslim} {Bias} in {Large} {Language} {Models}}, isbn = {978-1-4503-8473-5}, url = {https://dl.acm.org/doi/10.1145/3461702.3462624}, doi = {10.1145/3461702.3462624}, abstract = {It has been observed that large-scale language models capture undesirable societal biases, e.g. relating to race and gender; yet religious bias has been relatively unexplored. We demonstrate that GPT-3, a state-of-the-art contextual language model, captures persistent Muslim-violence bias. We probe GPT-3 in various ways, including prompt completion, analogical reasoning, and story generation, to understand this anti-Muslim bias, demonstrating that it appears consistently and creatively in different uses of the model and that it is severe even compared to biases about other religious groups. For instance, “Muslim" is analogized to “terrorist" in 23\% of test cases, while “Jewish" is mapped to its most common stereotype, “money," in 5\% of test cases. We quantify the positive distraction needed to overcome this bias with adversarial text prompts, and find that use of the most positive 6 adjectives reduces violent completions for “Muslims" from 66\% to 20\%, but which is still higher than for other religious groups.}, language = {en}, urldate = {2023-03-17}, booktitle = {Proceedings of the 2021 {AAAI}/{ACM} {Conference} on {AI}, {Ethics}, and {Society}}, publisher = {ACM}, author = {Abid, Abubakar and Farooqi, Maheen and Zou, James}, month = jul, year = {2021}, pages = {298--306}, } @misc{blodgett_language_2020, title = {Language ({Technology}) is {Power}: {A} {Critical} {Survey} of "{Bias}" in {NLP}}, shorttitle = {Language ({Technology}) is {Power}}, url = {http://arxiv.org/abs/2005.14050}, abstract = {We survey 146 papers analyzing “bias” in NLP systems, finding that their motivations are often vague, inconsistent, and lacking in normative reasoning, despite the fact that analyzing “bias” is an inherently normative process. We further find that these papers’ proposed quantitative techniques for measuring or mitigating “bias” are poorly matched to their motivations and do not engage with the relevant literature outside of NLP. Based on these findings, we describe the beginnings of a path forward by proposing three recommendations that should guide work analyzing “bias” in NLP systems. These recommendations rest on a greater recognition of the relationships between language and social hierarchies, encouraging researchers and practitioners to articulate their conceptualizations of “bias”—i.e., what kinds of system behaviors are harmful, in what ways, to whom, and why, as well as the normative reasoning underlying these statements—and to center work around the lived experiences of members of communities affected by NLP systems, while interrogating and reimagining the power relations between technologists and such communities.}, language = {en}, urldate = {2023-03-19}, publisher = {arXiv}, author = {Blodgett, Su Lin and Barocas, Solon and Daumé III, Hal and Wallach, Hanna}, month = may, year = {2020}, note = {arXiv:2005.14050 [cs]}, keywords = {Computer Science - Computers and Society, Computer Science - Computation and Language}, } @misc{barikeri_redditbias_2021, title = {{RedditBias}: {A} {Real}-{World} {Resource} for {Bias} {Evaluation} and {Debiasing} of {Conversational} {Language} {Models}}, shorttitle = {{RedditBias}}, url = {http://arxiv.org/abs/2106.03521}, abstract = {Text representation models are prone to exhibit a range of societal biases, reflecting the noncontrolled and biased nature of the underlying pretraining data, which consequently leads to severe ethical issues and even bias amplification. Recent work has predominantly focused on measuring and mitigating bias in pretrained language models. Surprisingly, the landscape of bias measurements and mitigation resources and methods for conversational language models is still very scarce: it is limited to only a few types of bias, artificially constructed resources, and completely ignores the impact that debiasing methods may have on the final performance in dialog tasks, e.g., conversational response generation. In this work, we present REDDITBIAS, the first conversational data set grounded in the actual human conversations from Reddit, allowing for bias measurement and mitigation across four important bias dimensions: gender, race, religion, and queerness. Further, we develop an evaluation framework which simultaneously 1) measures bias on the developed REDDITBIAS resource, and 2) evaluates model capability in dialog tasks after model debiasing. We use the evaluation framework to benchmark the widely used conversational DialoGPT model along with the adaptations of four debiasing methods. Our results indicate that DialoGPT is biased with respect to religious groups and that some debiasing techniques can remove this bias while preserving downstream task performance.}, language = {en}, urldate = {2023-05-24}, publisher = {arXiv}, author = {Barikeri, Soumya and Lauscher, Anne and Vulić, Ivan and Glavaš, Goran}, month = jun, year = {2021}, note = {arXiv:2106.03521 [cs]}, keywords = {Computer Science - Computation and Language}, } @article{lee_exploring_2019, title = {Exploring {Social} {Bias} in {Chatbots} using {Stereotype} {Knowledge}}, abstract = {Exploring social bias in chatbots is an important, yet relatively unexplored problem. In this paper, we propose an approach to understand social bias in chatbots by leveraging stereotype knowledge. It allows interesting comparison of bias between chatbots and humans, and provides intuitive analysis of existing chatbots by borrowing the finer-grain concepts of sexism and racism.}, language = {en}, author = {Lee, Nayeon and Madotto, Andrea and Fung, Pascale}, year = {2019}, } @inproceedings{feine2020gender, title={Gender bias in chatbot design}, author={Feine, Jasper and Gnewuch, Ulrich and Morana, Stefan and Maedche, Alexander}, booktitle={Chatbot Research and Design: Third International Workshop, CONVERSATIONS 2019, Amsterdam, The Netherlands, November 19--20, 2019, Revised Selected Papers 3}, pages={79--93}, year={2020}, organization={Springer} } @article{silva2019algorithms, title={Algorithms, platforms, and ethnic bias}, author={Silva, Selena and Kenney, Martin}, journal={Communications of the ACM}, volume={62}, number={11}, pages={37--39}, year={2019}, publisher={ACM New York, NY, USA} } @book{lawrence2023hidden, title={Hidden in White Sight: How AI Empowers and Deepens Systemic Racism}, author={Lawrence, Calvin D}, year={2023}, publisher={CRC Press} } @article{solon_barocas_problem_2017, title = {The {Problem} {With} {Bias}: {Allocative} {Versus} {Representational} {Harms} in {Machine} {Learning}}, journal = {Proceedings of SIGCIS, Philadelphia, PA}, author = {{Solon Barocas} and {Kate Crawford} and {Aaron Shapiro} and {Hanna Wallach}}, year = {2017}, } @article{frick2022transformer, title={Fraunhofer SIT at CheckThat! 2022: ensemble similarity estimation for finding previously fact-checked claims}, author={Frick, Raphael Antonius and Vogel, Inna}, journal={Working Notes of CLEF}, year={2022} } @inproceedings{siddique2022transformer, title={Personalizing Task-oriented Dialog Systems via Zero-shot Generalizable Reward Function}, author={Siddique, AB and Maqbool, MH and Taywade, Kshitija and Foroosh, Hassan}, booktitle={Proceedings of the 31st ACM International Conference on Information \& Knowledge Management}, pages={1787--1797}, year={2022} } @article{iqbal2023lady-transformer, title={Lady and the Tramp Nextdoor: Online Manifestations of Real-World Inequalities in the Nextdoor Social Network}, author={Iqbal, Waleed and Ghafouri, Vahid and Tyson, Gareth and Suarez-Tangil, Guillermo and Castro, Ignacio}, journal={arXiv preprint arXiv:2304.05232}, year={2023} } @misc{shen2023chatgpt, title={In ChatGPT We Trust? Measuring and Characterizing the Reliability of ChatGPT}, author={Xinyue Shen and Zeyuan Chen and Michael Backes and Yang Zhang}, year={2023}, eprint={2304.08979}, archivePrefix={arXiv}, primaryClass={cs.CR} } @article{ali2023chatgpt, title={ChatGPT and Lacrimal Drainage Disorders: Performance and Scope of Improvement}, author={Ali, M. J.}, journal={Ophthalmic plastic and reconstructive surgery}, volume={39}, number={3}, pages={221--225}, year={2023}, doi={10.1097/IOP.0000000000002418}, } @misc{TheStar2023WokeAI, author = "Kevin Jiang", title = "What is 'woke AI' and why is Elon Musk reportedly building a chatbot to counter it?", howpublished = "TheStar", year = "2023", month = "March", day = "2", url = "https://www.thestar.com/business/2023/03/01/what-is-woke-ai-and-why-is-elon-musk-reportedly-building-a-chatbot-to-counter-it.html", note = "Accessed on Month Day, Year", } @misc{Verge2023WokeAI, author = "Vincent, James", title = "As conservatives criticize ‘woke AI,’ here are ChatGPT’s rules for answering culture war queries", howpublished = "The Verge", year = "2023", month = "February", day = "17", url = "https://www.theverge.com/2023/2/17/23603906/openai-chatgpt-woke-criticism-culture-war-rules", note = "Accessed on Month Day, Year", } @article{beck_managing_2019, title = {Managing conflict in online debate communities}, volume = {24}, shorttitle = {Managing {Conflict} in {Online} {Debate} {Communities}}, abstract = {Kialo is a novel peer production system focused on pro/con debate construction. Teams of moderators vet and accept claims submitted by writers. Moderators also edit and refactor debates as they grow. Thus, moderators play a critical role in cultivating and maintaining debates. Conflict between moderators is typical. It is a feature of argumentation and debate. However, not all conflict is productive. Conflict between moderators can undermine collaboration (by distracting from the task of managing debates) and drive attrition (by discouraging participation on the site altogether). Based on a ten-month participant observation on Kialo, we identify a common source of conflict between moderators: adversarial beliefs and values. Moderators are not neutral participants on Kialo. They take positions on debate topics. We suggest foregrounding these positions, which are potential sources of conflict, through interface design as a scalable solution to conflict management.}, language = {en}, number = {7}, urldate = {2023-06-02}, journal = {First Monday}, author = {Beck, Jordan and Neupane, Bikalpa and Carroll, John M.}, month = jun, year = {2019}, doi = {https://doi.org/10.5210/fm.v24i7.9585}, } @misc{pew-research, author = "{Pew Research Center—U.S. Politics \& Policy (blog)}", title = "Political Typology Quiz", howpublished = "Available online", year = "n.d.", url = "https://www.pewresearch.org/politics/quiz/political-typology/", } @misc{eightvaluestest, author = "{IDRlabs}", title = "8 Values Political Test", howpublished = "Available online", year = "n.d.", url = "https://www.idrlabs.com/8-values-political/test.php", } @misc{political-compass, author = "{The Political Compass}", title = "Political Compass Test", howpublished = "Available online", year = "n.d.", url = "https://www.politicalcompass.org/test", } @inproceedings{Zhou2020allsides, author = {Zhou, Xinyi and Mulay, Apurva and Ferrara, Emilio and Zafarani, Reza}, title = {ReCOVery: A Multimodal Repository for COVID-19 News Credibility Research}, year = {2020}, isbn = {9781450368599}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3340531.3412880}, doi = {10.1145/3340531.3412880}, abstract = {First identified in Wuhan, China, in December 2019, the outbreak of COVID-19 has been declared as a global emergency in January, and a pandemic in March 2020 by the World Health Organization (WHO). Along with this pandemic, we are also experiencing an "infodemic" of information with low credibility such as fake news and conspiracies. In this work, we present ReCOVery, a repository designed and constructed to facilitate research on combating such information regarding COVID-19. We first broadly search and investigate ~2,000 news publishers, from which 60 are identified with extreme [high or low] levels of credibility. By inheriting the credibility of the media on which they were published, a total of 2,029 news articles on coronavirus, published from January to May 2020, are collected in the repository, along with 140,820 tweets that reveal how these news articles have spread on the Twitter social network. The repository provides multimodal information of news articles on coronavirus, including textual, visual, temporal, and network information. The way that news credibility is obtained allows a trade-off between dataset scalability and label accuracy. Extensive experiments are conducted to present data statistics and distributions, as well as to provide baseline performances for predicting news credibility so that future methods can be compared. Our repository is available at http://coronavirus-fakenews.com.}, booktitle = {Proceedings of the 29th ACM International Conference on Information \& Knowledge Management}, pages = {3205–3212}, numpages = {8}, keywords = {fake news, infodemic, repository, social media, information credibility, coronavirus, multimodal, covid-19, pandemic}, location = {Virtual Event, Ireland}, series = {CIKM '20} } @inproceedings{Deb2019allsides, author = {Deb, Ashok and Luceri, Luca and Badaway, Adam and Ferrara, Emilio}, title = {Perils and Challenges of Social Media and Election Manipulation Analysis: The 2018 US Midterms}, year = {2019}, isbn = {9781450366755}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3308560.3316486}, doi = {10.1145/3308560.3316486}, abstract = {One of the hallmarks of a free and fair society is the ability to conduct a peaceful and seamless transfer of power from one leader to another. Democratically, this is measured in a citizen population’s trust in the electoral system of choosing a representative government. In view of the well documented issues of the 2016 US Presidential election, we conducted an in-depth analysis of the 2018 US Midterm elections looking specifically for voter fraud or suppression. The Midterm election occurs in the middle of a 4 year presidential term. For the 2018 midterms, 35 Senators and all the 435 seats in the House of Representatives were up for re-election, thus, every congressional district and practically every state had a federal election. In order to collect election related tweets, we analyzed Twitter during the month prior to, and the two weeks following, the November 6, 2018 election day. In a targeted analysis to detect statistical anomalies or election interference, we identified several biases that can lead to wrong conclusions. Specifically, we looked for divergence between actual voting outcomes and instances of the #ivoted hashtag on the election day. This analysis highlighted three states of concern: New York, California, and Texas. We repeated our analysis discarding malicious accounts, such as social bots. Upon further inspection and against a backdrop of collected general election-related tweets, we identified some confounding factors, such as population bias, or bot and political ideology inference, that can lead to false conclusions. We conclude by providing an in-depth discussion of the perils and challenges of using social media data to explore questions about election manipulation.}, booktitle = {Companion Proceedings of The 2019 World Wide Web Conference}, pages = {237–247}, numpages = {11}, keywords = {political elections, data science for society, social media}, location = {San Francisco, USA}, series = {WWW '19} } @inproceedings{Ye2019allsides, author = {Ye, Junting and Skiena, Steven}, title = {MediaRank: Computational Ranking of Online News Sources}, year = {2019}, isbn = {9781450362016}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3292500.3330709}, doi = {10.1145/3292500.3330709}, abstract = {In the recent political climate, the topic of news quality has drawn attention both from the public and the academic communities. The growing distrust of traditional news media makes it harder to find a common base of accepted truth. In this work, we design and build MediaRank (urlwww.media-rank.com ), a fully automated system to rank over 50,000 online news sources around the world. MediaRank collects and analyzes one million news webpages and two million related tweets everyday. We base our algorithmic analysis on four properties journalists have established to be associated with reporting quality: peer reputation, reporting bias/breadth, bottomline financial pressure, and popularity. Our major contributions of this paper include: (i) Open, interpretable quality rankings for over 50,000 of the world's major news sources. Our rankings are validated against 35 published news rankings, including French, German, Russian, and Spanish language sources. MediaRank scores correlate positively with 34 of 35 of these expert rankings. (ii) New computational methods for measuring influence and bottomline pressure. To the best of our knowledge, we are the first to study the large-scale news reporting citation graph in-depth. We also propose new ways to measure the aggressiveness of advertisements and identify social bots, establishing a connection between both types of bad behavior. (iii) Analyzing the effect of media source bias and significance. We prove that news sources cite others despite different political views in accord with quality measures. However, in four English-speaking countries (US, UK, Canada, and Australia), the highest ranking sources all disproportionately favor left-wing parties, even when the majority of news sources exhibited conservative slants.}, booktitle = {Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery \& Data Mining}, pages = {2469–2477}, numpages = {9}, keywords = {online news ranking, mediarank, news citation network/graph}, location = {Anchorage, AK, USA}, series = {KDD '19} } @inproceedings{Fourney2018gunningfog, author = {Fourney, Adam and Ringel Morris, Meredith and Ali, Abdullah and Vonessen, Laura}, title = {Assessing the Readability of Web Search Results for Searchers with Dyslexia}, year = {2018}, isbn = {9781450356572}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3209978.3210072}, doi = {10.1145/3209978.3210072}, abstract = {Standards organizations, (e.g., the World Wide Web Consortium), are placing increased importance on the cognitive accessibility of online systems, including web search. Previous work has shown an association between query-document relevance judgments, and query-independent assessments of document readability. In this paper, we study the lexical and aesthetic features of web documents that may underlie this relationship. Leveraging a data set consisting of relevance and readability judgments for 200 web pages as assessed by 174 adults with dyslexia and 172 adults without dyslexia, we answer the following research questions: (1) Which web page features are most associated with readability? (2) To what extent are these features also associated with relevance? And, (3) are any features associated with the differences in readability/relevance judgments observed between dyslexic and non-dyslexic populations? Our findings have implications for improving the cognitive accessibility of search systems and web documents.}, booktitle = {The 41st International ACM SIGIR Conference on Research \& Development in Information Retrieval}, pages = {1069–1072}, numpages = {4}, keywords = {dyslexia, web search, readability}, location = {Ann Arbor, MI, USA}, series = {SIGIR '18} } @article{Bothun2022gunningfog, title = {Readability of COVID-19 vaccine information for the general public}, journal = {Vaccine}, volume = {40}, number = {25}, pages = {3466-3469}, year = {2022}, issn = {0264-410X}, doi = {https://doi.org/10.1016/j.vaccine.2022.04.096}, url = {https://www.sciencedirect.com/science/article/pii/S0264410X22005461}, author = {Luke S. Bothun and Scott E. Feeder and Gregory A. Poland}, keywords = {Research, Reading, Research Subjects, Consent Forms, Clinical Trial, Coronavirus Infections, COVID-19, Vaccine, Comprehension}, abstract = {Background More than 130 million individuals in the United States have now received at least one dose of a COVID-19 vaccine. Currently, all adults in the Unites States now have access to one of three COVID-19 vaccines. As part of the vaccination procedure, Emergency Use Authorization (EUA) fact sheets, which contain information regarding the vaccine, are provided. The purpose of this study was to analyze the ease of reading (i.e., readability) of the EUA-approved fact sheets for the vaccines currently available in the United States, the V-Safe adverse event survey script, and the Centers for Disease Control and Prevention (CDC) website information on COVID-19 vaccines designed for the general public in the United States. Methods We acquired the Pfizer, Moderna, and Janssen EUA fact sheets, as well as the V-Safe survey script and the CDC website information regarding COVID-19 vaccines. These documents were analyzed for their complexity regarding the following readability factors: average length of paragraphs, sentences, and words; font size and style; use of passive voice; the Gunning-Fog index; the Flesch Reading Ease index; and the Flesch-Kincaid Grade Level index. Results Only the V-Safe adverse-event survey script met readability standards for adequate comprehension. The mean readability scores of the EUA fact sheets and the CDC website were as follows: Flesch Reading Ease score (44.35 avg); Flesch-Kincaid Grade Level (10.48 avg); and Gunning-Fog index (11.8 avg).These scores indicate that at least a 10th-grade level education would be required to understand these reading materials. Conclusion The average person in the United States would have difficulty understanding the information provided in the EUA fact sheets and CDC COVID-19 vaccine website documents; however, the V-Safe survey was written at an adequate reading level. To ensure that the general public fully understands information regarding COVID-19 vaccines, greater care and effort should be given to the development of simplified information material.} } @article{Suleiman2016gunningfog, author = {Ahna Ballonoff Suleiman and Jessica S. Lin and Norman A. Constantine}, title = {Readability of Educational Materials to Support Parent Sexual Communication With Their Children and Adolescents}, journal = {Journal of Health Communication}, volume = {21}, number = {5}, pages = {534-543}, year = {2016}, publisher = {Taylor \& Francis}, doi = {10.1080/10810730.2015.1103334}, note ={PMID: 27116292}, URL = { https://doi.org/10.1080/10810730.2015.1103334 }, eprint = { https://doi.org/10.1080/10810730.2015.1103334 } } | es |
