Twitter Algorithm Amplifies Politically Right-Leaning Content: Study

The analysis result suggested that in six out of seven countries the amplification algorithm highly favoured the tweets from politically right-leaning sources.

Representational Image (Photo: Getty Images)

New Delhi: A recent study suggests that the algorithm used by the micro-blogging platform Twitter amplifies politically right-leaning content more than the left-leaning content. The algorithm which is used for personalization of the content according to the user works for amplification of political content unequally, reports news agency ANI.

The study was conducted by Cambridge University professor Ferenc Huszar and his colleagues, who quantified Twitter’s algorithm. Professor Huszar and his colleagues studied a randomised control sample of nearly two million daily active Twitter users chosen by the platform in reverse-chronological order without the personalised feature, and a treatment group representing four percent of all the users with personalisation feature on their timelines.

The team further analysed the algorithmic amplification effect on the tweets made by 3,634 elected political leaders from seven countries that are highly represented on Twitter. Added to this, they also analysed the similar impact on 6.2 million political news articles shared in the USA. 

The analysis result suggested that in six out of seven countries the amplification algorithm highly favoured the tweets from politically right-leaning sources. A similar trend was seen when it came to right-leaning US news services.

However, contrary to popular belief, far left or far right content was not as significantly amplified as the moderate content. The results collected by this study will be a contribution to the analysis of personalisation algorithms, as per the ANI report.