• lemmyvore@feddit.nl
    link
    fedilink
    English
    arrow-up
    0
    ·
    3 months ago

    In polls the demographics of the participants are selected so they are statistically significant.

    When you poll visitors to a website there’s no telling what biases might be at work. For example, Arch is a popular distribution so its prevalence in the results could simply be an indicator of that.

    Last but not least, correlation does not necessarily mean causation, and that goes especially for an uncontrolled sample. There’s a famous example that says “100% of the people who drink water die after that”.