Twitter's personalization algorithms amplify more right-leaning content than left-leaning, while "strong partisan bias" in news stories can boost amplification, but the algorithm does not favor political extremes on both sides more than moderate beliefs, according to an analysis of millions of tweets by researchers at Twitter.Social media platforms such as Facebook have long faced scrutiny for their roles in political discourse in this country, and researchers at Twitter sought to better understand their role. To do this, researchers split their study into two parts: The first part examined tweets from elected officials in Canada, France, Germany, Japan, Spain, the United Kingdom, and the US from April to August 2020. The second part examined how Twitter's two timeline algorithms (the Home timeline that prioritizes tweets that Twitter thinks you want to see and "Latest tweets," which shows tweets in reverse chronological order) amplified political content."In six out of seven countries—all but Germany—tweets posted by accounts from the political right receive more algorithmic amplification than the political left when studied as a group," Rumman Chowdhury, Director of Software Engineering, and Luca Belli, Staff Machine Learning Researcher, write in a blog post. But "group effects did not translate to individual effects. In other words, since party affiliation or ideology is not a factor our systems consider when recommending content, two individuals in the same political party would not necessarily see the same amplification," they wrote.Overall, tweets about political content from elected officials, regardless of party or whether the party is in power, do see algorithmic amplification when compared to political content on the "reverse chronological timeline," Twitter says. Since content from elected officials makes up a small portion of all political content on Twitter, researchers then examined algorithmic amplification of news outlets. "Content from US media outlets with a strong right-leaning bias are amplified marginally more than content from left-leaning sources," the report finds. Meanwhile, "some findings point at the possibility that strong partisan bias in news reporting is associated with higher amplification," it adds. But that could go either way, and "does not imply the promotion of extreme political ideology."As the report notes, a media org's political affiliation was determined by independent third-party sources (AllSides and AdFontes), which assigned a five-point score (with five being more right-wing) to each media source linked to in the tweets examined by researchers.Why does right-leaning content have a slight edge? Jury's still out. "Recent arguments that different political parties pursue different strategies on Twitter may provide an explanation as to why these disparities exist. However, understanding the precise causal mechanism that drives amplification invites further study that we hope our work initiates," Twitter says.While algorithmic content curation of political content was the focus of this study, this same methodology could be used to provide insights into misinformation, hate speech, and abusive content, the report suggests.
Wir nutzen Cookies auf unserer Website. Einige von ihnen sind essenziell für den Betrieb der Seite, während andere uns helfen, diese Website und die Nutzererfahrung zu verbessern (Tracking Cookies). Sie können selbst entscheiden, ob Sie die Cookies zulassen möchten. Bitte beachten Sie, dass bei einer Ablehnung womöglich nicht mehr alle Funktionalitäten der Seite zur Verfügung stehen.
Twitter's personalization algorithms amplify more right-leaning content than left-leaning, while "strong partisan bias" in news stories can boost amplification, but the algorithm does not favor political extremes on both sides more than moderate beliefs, according to an analysis of millions of tweets by researchers at Twitter.Social media platforms such as Facebook have long faced scrutiny for their roles in political discourse in this country, and researchers at Twitter sought to better understand their role. To do this, researchers split their study into two parts: The first part examined tweets from elected officials in Canada, France, Germany, Japan, Spain, the United Kingdom, and the US from April to August 2020. The second part examined how Twitter's two timeline algorithms (the Home timeline that prioritizes tweets that Twitter thinks you want to see and "Latest tweets," which shows tweets in reverse chronological order) amplified political content."In six out of seven countries—all but Germany—tweets posted by accounts from the political right receive more algorithmic amplification than the political left when studied as a group," Rumman Chowdhury, Director of Software Engineering, and Luca Belli, Staff Machine Learning Researcher, write in a blog post. But "group effects did not translate to individual effects. In other words, since party affiliation or ideology is not a factor our systems consider when recommending content, two individuals in the same political party would not necessarily see the same amplification," they wrote.Overall, tweets about political content from elected officials, regardless of party or whether the party is in power, do see algorithmic amplification when compared to political content on the "reverse chronological timeline," Twitter says. Since content from elected officials makes up a small portion of all political content on Twitter, researchers then examined algorithmic amplification of news outlets. "Content from US media outlets with a strong right-leaning bias are amplified marginally more than content from left-leaning sources," the report finds. Meanwhile, "some findings point at the possibility that strong partisan bias in news reporting is associated with higher amplification," it adds. But that could go either way, and "does not imply the promotion of extreme political ideology."As the report notes, a media org's political affiliation was determined by independent third-party sources (AllSides and AdFontes), which assigned a five-point score (with five being more right-wing) to each media source linked to in the tweets examined by researchers.Why does right-leaning content have a slight edge? Jury's still out. "Recent arguments that different political parties pursue different strategies on Twitter may provide an explanation as to why these disparities exist. However, understanding the precise causal mechanism that drives amplification invites further study that we hope our work initiates," Twitter says.While algorithmic content curation of political content was the focus of this study, this same methodology could be used to provide insights into misinformation, hate speech, and abusive content, the report suggests.