Algorithm Audit: Why, What, and How?

Aragona, Biagio

doi:10.4324/9781003080381

Cited by 7 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This invokes the general theme of data as a social construct and not as an object in its own right. It applies especially to algorithmic processes, which are seemingly "objective" but certainly lack neutrality because they respond to predetermined logics (Aragona, 2021).…”

Section: The Problematic Steps Of Digital Research: Collecting and In...mentioning

confidence: 99%

To scrape or not to scrape, this is dilemma. The post-API scenario and implications on digital research

Trezza

2023

Front. Sociol.

View full text Add to dashboard Cite

IntroductionThis article aims to investigate the potential impact of restricted social data access on digital research practices. The 2018 Cambridge Analytica scandal exposed the exploitation of Facebook user data for speculative purposes and led to the end of the so-called “Data Golden Age,” characterized by free access to social media user data. As a result, many social platforms have limited or entirely banned data access. This policy shift, referred to as the “APIcalypse,” has revolutionized digital research methods.MethodsTo address the impact of this policy shift on digital research, a non-probabilistic sample of Italian researchers was surveyed and the responses were analyzed. The survey was designed to explore how constraints on digital data access have altered research practices, whether we are truly in a post-API era with a radical change in data scraping strategies, and what shared and sustainable solutions can be identified for the post-API scenario.ResultsThe findings highlight how limits on social data access have not yet created a “post-Api” scenario as expected, but it is turning research practices upside down, positively and negatively. On the positive side, because researchers are experimenting with innovative forms of scraping. Negatively, because there could be a “mass migration” to the few platforms that freely grant their APIs, with critical consequences for the quality of research.DiscussionThe closure of many social media APIs has not opened up a post-API world, but has worsened the conditions of making research, which is increasingly oriented to “easy-data” environments such as Twitter. This should prompt digital researchers to make a self-reflexive effort to diversify research platforms and especially to act ethically with user data. It would also be important for the scientific world and large platforms to enter into understandings for open and conscious sharing of data in the name of scientific progress.

show abstract

Section: The Problematic Steps Of Digital Research: Collecting and In...mentioning

confidence: 99%

To scrape or not to scrape, this is dilemma. The post-API scenario and implications on digital research

Trezza

2023

Front. Sociol.

View full text Add to dashboard Cite

show abstract

“…In this sense, digital media literacy is crucial to tackle extremism on social media. People are often unaware of how social media algorithms work to direct their exposure to news (Aragona, 2021) and to expose them to extreme content. Although digital media literacy is, of course, not a panacea in solving extremism on social media, it can still prevent some people from falling into extremism.…”

Section: Preventing Polarizationmentioning

confidence: 99%

The banality of extremism: The role of group dynamics and communication of norms in polarization on January 6.

Swol¹,

Lee²,

Hutchins³

2022

Group Dynamics: Theory, Research, and Practice

View full text Add to dashboard Cite

Many of the individuals charged in the January 6, 2021, attack on the U.S. Capitol did not have ties to far-right extremist groups, leading some pundits and researchers to speculate that extremism has become mainstream and much more common in the current political environment. In this article, we discuss events before and during January 6 that affected extremism and pushed people toward polarization. Specifically, we discuss the role of the media, especially social media, in developing like-minded content sharing and polarizing, extremist group norms. The role of groups to drive people toward protest is discussed within research on social movements. The role of identity and acceptability of prejudice toward outgroup political members and how this identity fed into increasing extremist rhetoric are also discussed. Then, we discuss conditions of the day of January 6, including establishment of norms through rhetoric and the social proof of extremist action, and also, the easy exit of those who could leave an increasingly escalating situation. Finally, we address the implications for the current political climate, including the potential for increased violence and extremism. Highlights and Implications• This article seeks to answer the question: What drove people normally not associated with far-right extremists to take part in the January 6 attack on the U.S. Capitol?• This article highlights the role of social media in creating homogeneous networks, where ideologically like-minded individuals shared emotions, concerns, and grievances.• The role of identity and affective polarization in driving outgroup animosity and affectively polarizing rhetoric is discussed.• Events during January 6 that contributed to violence include polarizing rhetoric, easy exit for dissenters, escalating social proof, and development of more extreme norms.• Implications of increasing political violence are discussed, as the events of January 6 have further polarized the public in the United States.

show abstract

“…It is important to make the results of such audits available to the broader public to increase social awareness about risks inherent in the use of algorithms by citizens, by policymakers, and the institutions they belong to (Aragona, 2021). The challenge here is twofold.…”

Section: Future Directionsmentioning

confidence: 99%

“…Search engine providers, such as Google or Bing, are under academic scrutiny due to their role as gatekeepers: they decide on the relevance of content by filtering and ranking sources of information available on the web (Laidlaw, 2010;Schulz et al, 2005;Wallace, 2018). The output of these decisions shapes our social reality (Just and Latzer, 2017;Noble, 2018) and their implications for citizens and institutions need to be carefully reflected (Aragona, 2021;Halavais, 2018;Thomas et al, 2018). However, the algorithmic choices are not openly accessible (Kitchin, 2017;Paßmann and Boersma, 2017), making independent audits of the outputs crucial.…”

Section: Introductionmentioning

confidence: 99%

Representativeness and face-ism: Gender bias in image search

Ulloa

Richter

Makhortykh

et al. 2022

New Media & Society

View full text Add to dashboard Cite

Implicit and explicit gender biases in media representations of individuals have long existed. Women are less likely to be represented in gender-neutral media content (representation bias), and their face-to-body ratio in images is often lower (face-ism bias). In this article, we look at representativeness and face-ism in search engine image results. We systematically queried four search engines (Google, Bing, Baidu, Yandex) from three locations, using two browsers and in two waves, with gender-neutral (person, intelligent person) and gendered (woman, intelligent woman, man, intelligent man) terminology, accessing the top 100 image results. We employed automatic identification for the individual’s gender expression (female/male) and the calculation of the face-to-body ratio of individuals depicted. We find that, as in other forms of media, search engine images perpetuate biases to the detriment of women, confirming the existence of the representation and face-ism biases. In-depth algorithmic debiasing with a specific focus on gender bias is overdue.

show abstract

Algorithm Audit: Why, What, and How?

Cited by 7 publications

References 0 publications

To scrape or not to scrape, this is dilemma. The post-API scenario and implications on digital research

To scrape or not to scrape, this is dilemma. The post-API scenario and implications on digital research

The banality of extremism: The role of group dynamics and communication of norms in polarization on January 6.

Representativeness and face-ism: Gender bias in image search

Contact Info

Product

Resources

About