Abstract
Through mixed-methods and a sociological theoretical framework from science and technology studies, this research addresses three standpoints in online dating: Graphical User Interfaces, Users and Developers. It is argued that together, these actants (i.e., humans and objects with agency) learn how to build a type of affective and algorithmic communication: dating app conventions (e.g., functionalities, variables, metrics) serve in measuring interactions and profiling personal preferences. At the same time, they establish a trial-error dynamic between humans and the machine for identifying and reviewing affordances that matter to each of them. The results map the dating app market through interface structures and shed light on how coding and app usages enable -or not- finding a date or a partner.
Website
Visit the dedicated website to this project.
Contact
Principal Investigator
Constance de Saint Laurent
Affiliations
University of Bologna and EPFL
Funding
Swiss National Science Foundation
Abstract
The aim of this project is to look at the social dynamics of online interactions, particularly in political discussions, and the effects of online misinformation and conspiracy theories on how we interact with others. It will investigate whether the use of rumours, conspiracy theories, and fabricated ‘facts’ is correlated with a tendency to dehumanise others and encourage violence. Indeed, the existing literature in social sciences on online misinformation indicates that it tends to be the prerogative of politically extreme and rather violent groups – violent at least in words.
In order to answer this question, this project focuses on the platform reddit, known for hosting debates with a general freedom of tone and content and yet reaching a quite wide audience in terms of political orientation. In particular, it aims to look at posts sharing online newspaper articles and the ensuing discussions, focusing the articles coming from websites varying in quality (reliable sources, websites with a tendency to spread misinformation, and clickbaits) and political orientation (centre, left leaning, and right leaning) for the period 2016-2019.
The analysis explores the effects of misinformation, rumours and conspiracy theories on conversational patterns, tone, and most importantly perspective taking. Indeed, this psychological process is at the heart of how we represent others, both present and absent, and therefore has the potential to shed light on the role of « alternative facts » and conspiracies on interactions with others and on the dehumanisation of other social groups.
This project should allow for a better understanding of the effects of « post-truth » – often presented as an era where feelings matter more than facts – on social interactions, and thus some of the long terms social consequences it may have. It is also an occasion to explore whether there is indeed a link between the decline of trust in expertise and science and the rise of far-right discourses. Finally, it will give us the opportunity to observe and analyse the dynamics of social interactions on social media, an understudied environment compared to its current prominence in social and political life.
Participant information
What is this project about?
This project looks at how news articles are discussed on reddit. Its aim is to compare the discussions below posts that link to trustworthy news sources with posts linking to unreliable or fake news websites. It also looks at the differences between biased news sources, may they be left-leaning or right-leaning, and more neutral ones.
How do you decide if a website is trustworthy or biased?
All the evaluations are based on ratings shared by NGOs and consortiums of journalists who have specialised in news source evaluations. In particular, the ratings of AllSides, the Unreliable News Site index, NewsGuard, FactCheck, Fake News Codex, OpenSources, PolitiFact, and Snopes were used.
Is my reddit account concerned?
If you posted a link to one of the websites listed at the end of this document between January 2016 and October 2019 or commented on one, then yes, chances are some of your data has been collected.
What data is being collected?
The main information collected is the content (text or link) of your post or comment, the time at which it was made, whether it has been edited, and the number of upvotes/downvotes. Other information concerns the subreddit in which it was posted. Your user ID is collected in order to identify all the comments made by a single user on a post (but not across posts), but it is not saved. In practice, it means that your user ID – which is different from your public username – is stored for a few minutes the time for all the comments made on a post to be collected. But it is replaced by a random number before we store it, and no directly identifying information is stored.
Why have I not been informed that my data was being collected?
Unfortunately, it is not possible to contact all of the users concerned prior to data collection, as we do not know beforehand who is concerned. We would thus have to contact the whole of reddit, which is technically not possible. Informing you afterwards would involve collecting and storing private information – such as your username – that would represent a disproportionate privacy risk. While researchers try as much as possible to inform research participants, it is part of reddit’s privacy policy to let third parties like researchers access the public part of your data. If you have any concern about this project in particular, we would be happy to answer your questions (see contact below). However, for general concerns about the use of your data on reddit, we invite you contact them directly, as they will be able to answer your questions.
What will you do with my data?
Your data will be stored on a password protected hard drive and on the secure platform Switch until December 2024, when it will be destroyed. It will be used in analyses looking at the types of topics discussed, the vocabulary used, and the reactions it led to from other users. The data in itself will only be accessed by the principal investigator. The results will be published in scientific articles and shared in conference presentations and university lectures. It will sometimes involve showing quotes (the content of your posts and comments). However, you will not be identifiable: the content of the quote will be altered so that it cannot be googled to find your account, and no other identifiable information will be shared alongside the quote.
Can I opt out?
Yes, you can opt out by sending an email to this e-mail address. Bear in mind, however, that because your username has been removed from the database, I will first need to collect all your posts and comments in order to delete them from our records. Alternatively, you can send the link to specific posts and comments you made and would like to see removed.
What if I have any question or concern?
You can send any question or concern to this e-mail address.
Can I see the results?
All the publications of the results will be shared here
List of selected websites
abcnews.go.com – activistpost.com – aljazeera.com – alternet.org – americanthinker.com – apnews.com – bbc.com – beforeitsnews.com – bloomberg.com – bostonglobe.com – breitbart.com – c-span.org – cbsnews.com – charismanews.com – cnn.com – cnsnews.com – collective-evolution.com – commondreams.org – conservativedailypost.com – csmonitor.com – dailycaller.com – dailykos.com – dailywire.com – democracynow.org – drudgereport.com – economist.com – express.co.uk – fivethirtyeight.com – foreignaffairs.com – foxnews.com – freebeacon.com – gellerreport.com – huffingtonpost.com – humansarefree.com – ilovemyfreedom.org – independent.co.uk – infowars.com – jacobinmag.com – jihadwatch.org – judicialwatch.org – latino.foxnews.com – lifezette.com – mediaite.com – mediamatters.org – motherjones.com – msnbc.com – nationalreview.com – nbcnews.com – neonnettle.com – news.yahoo.com – newsmax.com – newsweek.com – newyorker.com – nowtheendbegins.com – npr.org – nymag.com – nytimes.com – palmerreport.com – patriotpost.us – pjmedia.com – politico.com – politicususa.com – politifact.com – pri.org – rawstory.com – redstate.com – redstatewatcher.com – reuters.com – rollingstone.com – salon.com – shareblue.com – slate.com – spectator.org – teaparty.org – telegraph.co.uk – theamericanmirror.com – theatlantic.com – theblaze.com – theconservativetreehouse.com – thedailybeast.com – theepochtimes.com – thefederalist.com – thefreethoughtproject.com – thegatewaypundit.com – theguardian.com – thehill.com – theintercept.com – thelibertarianrepublic.com – themindunleashed.com – thenation.com – thenewamerican.com – thepoliticalinsider.com – therightscoop.com – theverge.com – theweek.com – thinkprogress.org – townhall.com – trueactivist.com – truepundit.com – truthdig.com – twitchy.com – usatoday.com – usnews.com – vanityfair.com – vice.com – vox.com – washingtonpost.com – washingtontimes.com – weeklystandard.com – westernjournalism.com – wnd.com – yahoo.com – yournewswire.com
This study has been approved by the EPFL Research Ethics Committee
(HREC No. 064-2019)