Sat May 5, 2018, 09:33 PM
Loki Liesmith (4,323 posts)
Why do so few users here not understand that single polls don't mean much?
It’s the poll average that matters. Don’t people get that yet?
|
8 replies, 1419 views
![]() |
Author | Time | Post |
![]() |
Loki Liesmith | May 2018 | OP |
former9thward | May 2018 | #1 | |
Loki Liesmith | May 2018 | #4 | |
Hoyt | May 2018 | #2 | |
pnwmom | May 2018 | #3 | |
Loki Liesmith | May 2018 | #5 | |
Loki Liesmith | May 2018 | #6 | |
RockRaven | May 2018 | #7 | |
Loki Liesmith | May 2018 | #8 |
Response to Loki Liesmith (Original post)
Sat May 5, 2018, 09:41 PM
former9thward (23,515 posts)
1. Maybe you should provide context.
It helps...
|
Response to former9thward (Reply #1)
Sat May 5, 2018, 09:49 PM
Loki Liesmith (4,323 posts)
4. Just see a lot of panic
Over a single “poll” (really a Reuters online probability sample) that shows Trump at 49%.
This at a time when 2 other simultaneous polls have Trump down 15 and 9 (Pew and Monmouth respectively). Given the statistical standards we use about 1 of every 20 polls should show bizarre or surprising results. I’m not going to call out a single poster but it’s discouraging to see people who have been here longer than me freaking out. Haven’t we learned the game yet? |
Response to Loki Liesmith (Original post)
Sat May 5, 2018, 09:43 PM
Hoyt (47,825 posts)
2. I think most do, but it is disturbing when object of poll should have a very low approval rating.
And many who disapprove do so only because the object of the poll isn’t deporting enough immigrants, dropping more bombs, shooting more innocent minorities, destroying safety net even faster, and worse.
|
Response to Loki Liesmith (Original post)
Sat May 5, 2018, 09:46 PM
pnwmom (104,013 posts)
3. And online, self-selected polls are always unreliable, no matter how much
they try to dress themselves up as real polls.
If they don't include a "margin of error," then they're a pseudo-poll, a non-scientific poll and no one should be posting them here. |
Response to pnwmom (Reply #3)
Sat May 5, 2018, 09:50 PM
Loki Liesmith (4,323 posts)
5. True
They’ve gotten better but still not proven methodologies.
|
Response to pnwmom (Reply #3)
Sat May 5, 2018, 09:51 PM
Loki Liesmith (4,323 posts)
6. I'm cool with them posting those results
But they should put them in proper context...
|
Response to Loki Liesmith (Original post)
Sat May 5, 2018, 10:08 PM
RockRaven (8,084 posts)
7. Innumeracy is widespread. nt
Response to Loki Liesmith (Original post)
Sat May 5, 2018, 10:19 PM
Loki Liesmith (4,323 posts)
8. We are overdue for a CNN poll
Bet we see one tomorrow.
|