Facebook's Algorithm Is ‘Influential' but Doesn't Necessarily Change Beliefs, Researchers Say

Facebook and Instagram's algorithms, which determine what millions of people see in social media, have been the target of a lot of criticism.

You can find out more about it here.

,

The following are some examples of the work of

You can also find out more about the following:

Regulators

Since years. Many people have called for algorithms to be removed to stop the spread of misinformation.

Four new studies, including one which examined data from 208,000,000 Americans who used Facebook during the 2020 Presidential Election, complicate this narrative.

Researchers from Princeton, New York University and the University of Texas found in their papers that removing certain key functions of social platform algorithms had no measurable effect on people's beliefs. Researchers found that in one experiment involving Facebook's algorithm the people's ability to share posts decreased when it was removed.

According to a second study, consumption of political information on Facebook and Instagram is highly segregated based on ideology. The research revealed that 97 percent of people who clicked on links to 'untrustworthy news' stories in the apps during 2020's election were conservative and viewed mostly right-wing material.

These studies, published in Science and Nature journals, paint a nuanced and contradictory picture of Americans' use of and reactions to two of the largest social media platforms on the planet. These contradictory results suggest that it may take many years to understand the role of social media in shaping discourse.

These papers were also notable for the number of Facebook and Instagram users included, and the fact that the researchers obtained the data and formulated the experiments in collaboration with Meta, the app's owner. These studies are part of a series that includes 16 peer-reviewed articles. Social media studies in the past have been based mainly on publicly-available information or on a small number of users who had information that they'scraped' or downloaded from the internet.

Talia Stroud is the founder and Director of the Center for Media Engagement at University of Texas at Austin. Joshua Tucker is a Professor and Co-Founder of the Center for Social Media and Politics at New York University. They led the project.

In an interview, Ms. Stroud stated that research revealed the 'complex social issues' we are dealing with and that there is likely to be 'no magic bullet' when it comes to social media's impact.

Katie Harbath said, "We need to be cautious about what we think is happening and what is actually happening." Katie Harbath was a former director of public policy at Meta, who left the company 2021. She said that the studies re-evaluated the 'assumed impact of social media'. She said that people's political preferences can be influenced by a variety of factors. 'Social media alone isn't to blame for our problems,' she added.

Meta is a term that describes a person.

It would be participating in the research by August 2020. The National Opinion Research Center, an independent agency at the University of Chicago that helped collect some of the data, spent $20,000,000 on the project. The company didn't pay the researchers even though some of their employees worked with them. Meta had the ability to reject data requests that would violate its users' privacy.

Michael Wagner, professor of mass communication at the University of Wisconsin Madison, an independent auditor for the project, stated that the work was not a good model for future research because it required the direct participation of Meta, which had all the data but only provided certain types to researchers. Researchers said that they were the final arbiters of the conclusions in their papers.

Nick Clegg is Meta's President of Global Affairs.

You can also read about the importance of this in our article

The studies revealed that 'there's little evidence' that Meta's key features alone can cause 'affective polarization' or have meaningful effects on outcomes. The findings will not settle the debate on social media and democracy, but he added that 'we expect and hope' they will help society understand these issues.

The papers are published at a time of turmoil in the social media world. This month,

Meta rolled Out Threads

Twitter is not the only social network. Elon Musk has made several changes to Twitter.

Renaming it X

Mastodon, YouTube, Reddit, and Discord are all thriving.

Bluesky

Appearing to gain some traction.

Meta has been trying to shift its focus from social apps and its immersive digital world in recent years.

Metaverse is the term used to describe the metaverse

Meta has experienced a lot of growth over the last 18 months.

Operating losses of more than $11 billion

Reality Labs, the division responsible for creating the metaverse.

Researchers have been raising questions for years about the algorithms that Facebook and Instagram use to determine what users see on their feeds. In 2021

Frances Haugen

Former Facebook employee, turned whistleblower, has further brought them to light. She gave lawmakers and the media thousands of documents from her company.

Facebook's algorithm was revealed by Congress

It was causing teenagers in Ethiopia to be exposed more to anorexia and 'literally fanned ethnic violence'.

Lawmakers include Senator Amy Klobuchar of Minnesota and Senator Cynthia Lummis of Wyoming.

Later introduced bills

Study or limit the algorithm. No one has passed.

In three of the four studies that were published on Thursday, Facebook users and Instagram users consented and provided their personal information to be used in order to take part. The company gave researchers anonymized data from 208 million Facebook users for the fourth study.

One study was entitled 'How does social media feed algorithm affect attitudes? In this research, which included over 23,000 Facebook and Instagram users, the researchers replaced the algorithm with reverse chronological feeds. This meant that people saw the latest posts first, instead of posts which were tailored to their preferences.

Researchers found that despite the change in feeds, people's political knowledge or 'polarization' did not change. The academics' surveys found that people didn't report changing their behavior, like signing more petitions online or attending more rallies after their feeds had been changed.

According to the study, it is alarming that a feed in reverse chronology increased the amount untrustworthy material people see.

According to a study of 208 million American Facebook Users during the 2020 elections, they were divided based on their political ideologies. Conservatives saw more misinformation compared to liberals.

According to research, conservatives tend to read more links to political news that are also read by other conservatives. More than 97 per cent of the news articles that were marked as false by third-party fact-checkers, were read by conservatives. Facebook Groups and Pages, which allow users to follow topics that interest them, shared links to more hyperpartisan articles.

The study found that Facebook Groups and Pages were a "very powerful curation machine" and "dissemination machine".

Researchers said that despite the fact that Facebook users are reading a low proportion of false articles, they still read fewer than all other news articles.

Researchers found in another study that the reduction of the content posted by 'like minded' connections on the feeds of 23,000 Facebook users did not have a measurable impact on the political or religious polarization.

The study's authors stated that 'these findings challenge popular narratives that blame social media echo chambers as the cause of the problems in contemporary American democracy'.

In a fourth experiment, which involved 27,000 Facebook users and Instagram users in a study, the people reported that their knowledge of current events had decreased when they were denied the ability to share posts. The paper concluded that removing the re-share option did not affect people's opinions or beliefs.

Researchers warned that the findings of their research were influenced by many factors. Some of the experiments could have been conducted right before the 2020 Presidential Election, which would mean that the users' political views were already cemented.

Some findings could be outdated. Meta, since the researchers began their work, has stopped displaying news content from publishers on users' Facebook and Instagram main news feeds. The company regularly adjusts and tweaks its algorithms in order to keep users interested.

Researchers said that they hoped their papers would spur more research in this field with the participation of other social media companies.

New York University's Tucker said, 'We hope the society will act through its policymakers to continue this type of research in the future.' This should be in the interest of society.