United KingdomA 2021 study by Stanford psychologist Michal Kosinski used facial recognition software to analyze over 1 million faces from the US, UK, and Canada, finding that algorithms could predict political orientation with 73% accuracy. This outperformed human judgment, which correctly identified political affiliation around 55% of the time. The research showed that subtle facial cues—such as head posture and expressions—could reliably indicate political leaning, even after excluding obvious demographic factors like race or age. The article then shares a personal anecdote about attending a Conservative political gathering, highlighting the contrast between the event's grandeur and the author's previous experiences with less formal, centrist political meetings.
Bias read (Right): The article frames the study's findings as evidence supporting a broader cultural narrative linking physical appearance with political ideology, particularly associating attractiveness with right-wing views. While the study itself is presented neutrally, the article emphasizes the implications of 'h
Why these scores (Factual 85 · Objective 65): Factuality is high as it describes a different study with similar findings, though it doesn't reference the Guardian study directly. Objectivity is lower due to the sensationalist title 'Why hot people are Right-wing' and the emphasis on the algorithm's success without acknowledging potential biases
