In a fascinating experiment called #WearthePants, women on LinkedIn tested the hypothesis that the platform’s algorithm was biased against them. Michelle, a product strategist, changed her gender to male and saw a significant increase in post impressions. This led to further investigations by other women like Marilynn Joyner, who experienced a 238% jump in impressions after changing her gender on her profile.
LinkedIn, however, denied any gender bias in its algorithm, stating that demographic information is not used to determine content visibility. Social algorithm experts suggest that while explicit sexism may not be the cause, implicit bias could be at play due to the intricate nature of algorithmic systems.
The #WearthePants experiment, initiated by entrepreneurs Cindy Gallop and Jane Evans, revealed disparities in post reach between men and women posting identical content. This raised concerns among women like Joyner, who called for LinkedIn to address any bias in its algorithm.
Researchers point out that popular language models are often trained on human-generated content, leading to biases based on societal norms. LinkedIn, like other platforms, keeps the specifics of its AI systems confidential, making it challenging to determine the exact reasons behind the disparities in post visibility.
While LinkedIn continues to refine its algorithm to provide a less biased experience, challenges remain in understanding how demographics, writing styles, and user behavior influence content visibility. The platform tests millions of posts to connect users with relevant opportunities, taking into account various signals to determine what appears in users’ feeds.
Despite efforts to improve algorithm transparency, content creators like Michelle and Shailvi Wakhulu express frustration over the fluctuating engagement levels on LinkedIn. While some users adapt to the new algorithm by focusing on specific topics, others struggle to maintain visibility and reach among their followers.
In conclusion, the complexities of algorithmic systems make it difficult to pinpoint the exact reasons behind disparities in post visibility on LinkedIn. Transparency and understanding of these systems remain crucial for content creators seeking to navigate the evolving landscape of social media algorithms.
