If YouTube algorithms radicalize users, the data doesn’t show it


We’ve all seen it it happened: Watch one video on YouTube and your recommendations change, as if Google’s algorithms think the theme of the video is your life’s passion. Suddenly, all the recommended videos – and probably many ads – are presented to you on this topic.

Basically, the results are comical. But it came to a constant flow of stories about how the process radicalized humans, sending them down ever deeper rabbit holes until their gaze was dominated by marginal ideas and conspiracy theories.

A new study released Monday considers whether these stories represent a bigger trend or are just a collection of anecdotes. Although the data cannot rule out the existence of radicalization online, they definitely suggest that this is not the most common experience. Instead, marginal ideas seem to be simply part of a larger community that is only getting stronger.

Big Data

Typically, the challenge of this kind of research is to collect data on people’s video-watching habits without these people knowing it — and potentially change their behavior. Researchers have circumvented this issue by obtaining data from Nielsen, which simply tracks what people are watching. People allow Nielsen to monitor their habits, and the company anonymizes the resulting data. For this study, the researchers collected data from over 300,000 viewers who together watched over 21 million videos on YouTube in the period from 2016 to the end of 2019.

Most of these videos had nothing to do with politics, so the authors used the literature to identify a large collection of channels that previous research had marked according to their political leanings, ranging from far left to centrist to far right. To that list, the researchers added a category they called “against waking up.” Although not always openly political, the growing collection of channels focuses on “opposing progressive social justice movements”. Although these channels tend to align with right-wing interests, video presenters often do not present ideas in this way.

Overall, the channels categorized by the researchers (just under 1,000 of them) accounted for only 3.3 percent of total video views during this period. And those who watched them stuck to one kind of content; if you started watching left-wing content in 2016, you will probably continue to watch it when the 2020 study period ends. In fact, based on the time spent per video, it is very likely that you will watch more of that content in 2020, perhaps as a product of controversy. Trump’s age.

(The exception to this is far-left content, which was viewed so rarely that in most cases it was impossible to single out statistically significant trends.)

Almost all types of off-edge content also recorded growth in this period, both in terms of total viewership and in terms of time spent watching videos on these channels (with the exception of far-left and far-right content). This finding suggests that at least some of the trends reflect the growing use of YouTube as a replacement for more traditional electronic media.


Since viewers mostly watched one type of content, it is easiest to imagine them as different groups. The researchers tracked the number of people belonging to each group, as well as the time they spent watching videos over a four-year period.

During that time, the main left was about as big as the other groups put together; he was followed by centrists. The main right and anti-awakening began the period at approximately the same level as the far right. But they all showed different trends. The total number of far-right viewers remained the same, but the time they spent watching the videos increased. In contrast, the total number of viewers of the mainstream right increased, but the time they spent watching did not differ much from that of the far right.

Anti-Awakening viewers showed the highest growth rate of all groups. By the end of the period, they spent more time watching videos than the centrists, even if their population remained smaller.

Source link


Please enter your comment!
Please enter your name here