1579
Comments (27)
sorted by:
You're viewing a single comment thread. View all comments, or full comment thread.
33
ThatGuyYouKnow 33 points ago +33 / -0

The world learned in 2016 that polls are useless, most or biased and all misrepresent the facts. I learned in a statistical analysis class that you can skew statistics to show anything you want just by changing the wording slightly.

14
nickybops [S] 14 points ago +14 / -0

Ive heard they can manipulate which type of Independents to call by their address. Im guessing calling independents in the big cities of a state will turnnout a higher number of left leaning independents then lets say if they were to call more rural living independents who may lean to the right more.

9
Watermelons 9 points ago +9 / -0

The bottom text says the data is weighted by things but even the unweighted total doesn't match. 566+248+336=1150 != 1200.

Either way, with the unweighted numbers, and assuming that the true proportions of D/R/I are 0.37, 0.33, and 0.30 respectively, the probability that you'd get 566/248/336 D/R/I in a true random sample is approximately 1.76*10^-19. or about 5 in 1 quintillion. In other words, they over-sampled D by a lot.

5
Crisis83 5 points ago +5 / -0

There are lies, damn lies and then there are statistics.

2
HockeyMom4Trump 2 points ago +2 / -0

100% of people surveyed agree with your response. At least I do, and I only surveyed myself, but that’s 100%