• Saleh@feddit.org
    link
    fedilink
    arrow-up
    1
    ·
    15 hours ago

    Thats a perfectly reasonable sample size and margin of error. The margin of error for a given confidence level and standard deviation is proportional to 1/sqrt(n) so in order to get the margin of error down from 4% points to 2% they would need to quadruple the number of people polled. However this is worthless if you cannot attribute for systematic errors with a better precision.