Posted on 08/14/2016 5:28:25 PM PDT by BlackFemaleArmyColonel
Ever since the Democratic Convention, the only poll results being reported by the liberal mainstream media show that Hillary Clinton is slowly pulling ahead of Donald Trump. Of course the same liberal mainstream media had been conducting a biased editing and reporting campaign to show Trump in a bad way and glorify Hillary.
They edited his comments about the Muslim parents of whose son was killed in action while serving in the US military. Trump stated about the mom not speaking, the liberal media intentionally edited out when Trump stated that that’s what others have written. They made it to look like those were Trump’s own words.
Now all of the controversy surrounding the way liberals have twisted Trump’s comment about people who value the Second Amendment. Hillary and even the Kennedy family have twisted it in such a way as declare Trump is inciting violence in some form. However, none of the same liberal media, nor Clinton nor the Kennedy family mention the fact that Hillary once stated that Obama could be assassinated.
The sad part is that so many people today believe everything they hear and see from the liberal mainstream media. They take the reports as gospel.
Did you know that in most of the polls, where they are reporting a Clinton lead, are conducted in such a way as to bias the outcome? Often, they only poll 100 up to 1,000 people. Not only are their sample sizes statistically small, but they often poll areas that are known to be predominately Democratic, hence, it only makes sense that Hillary is ahead.
But you probably have not heard about one poll that was recently taken where they polled 1,000 people from every state.
(Excerpt) Read more at constitution.com ...
How did they get the 1,000 in Rhode Island?
Sounds like a hoax.
I think you're right. I do not see any way 50,000 people were polled. That would mean in excess of 150,000 calls. Furthermore, it makes no sense to poll 1,000 people from small states like Rhode Island and Montana and weight them the same as 1,000 people from California.
Moreover, as someone else pointed out, increasing the sample size only increases accuracy by a small amount. A rational pollster would not poll 50,000 people.
If on election day, the actual turnout of voters is 33% for each party perfectly split, then this poll is perfect and 100% accurate.
The ratios matter, so if 40% of republicans come out and 50% of democrats actually come out and only 10% independents come out on election day, then this all goes out the window.
The good news is that the mainstream polls use the 2008 or 2012 ratios in the calculation for turnout, and in these years democrats came out and voted way higher than republicans.
This year, the primaries have indicated that republican and independent voters are way up and democrat voters are way down. So this is where the mainstream polls may be way off. They are expecting an Obama turnout for the old hag and we all know that people arent going to come out of the woodwork for her like they did for obama.
Did this poll concern itself with electoral college results or popular vote mechanics?
For the purposes of this discussion, don’t conflate the two.
CA....
Thanks for the screenshot!!!
http://www.freerepublic.com/focus/f-news/3459563/posts?page=39#39
Ping to screenshot of that gimmicked CNN poll !!
*
**
http://www.freerepublic.com/focus/f-news/3459563/posts?page=39#39
Ping to screenshot of that gimmicked CNN poll at # 39 !!
The poll is meaningless with the methodology used.
Trump is looking at a 50-State sweep that will make Reagan's look puny. He'll get over 60% of the vote with 4 choices on most state ballots.
Herre are the results of the five day NBC poll with over 136,00 votes.
http://www.poll-maker.com/results776652xBB154110-31#tab-2
True. A good poll has minimum 1500 respondents. After that is quality check on the first 1500. Anything less is weak.
It’s ridiculously implausible on it’s face and shouldn’t have been posted.
We can win this race but not with pie in the sky BS. This is a close race where the winner will get less than 50% of popular vote.
remember none of these are counting those that are dead vote.
Both polls may only have a slim margin of error, but they are only valid on the sample of people polled. If you polled 1000 African Americans you might get 80% Hillary 20% Trump with a 4% ME and 95%CI, but that would not accurately reflect the outcome of the vote because as you know that sample does not represent the electorate.
All polls, to be accurate, need to sample a grouping representative of the likely voting electorate. While it is impossible to predict who will turn out on election day, a poll would be closer to accurate if it reflected historical trends such as likely voters by party affiliation. For example if in 2012 4% more registered Democrats voted than Republicans, then a poll that sampled 10% more Democrats would not likely accurately project the outcome in November.
This is why they try to “weight polls”, but there are a number of different ways to look at the voters e.g. by race, sex, ethnicity, party affiliation, age and the like. The smaller the sample the less likely it will accurately reflect the electorate. A better poll would be larger, but it has to then be re-assembled by pulling in responses with a weighting that resembles the body politic. No method is perfect.
I live in Florida and they call here about three times a week. Different polls every time.
This is the one by P.G. farnsworth and its like pulling teeth to get to the internals which is everything no matter what kind of poll it is.
I posted this eight days ago on the tv talk show thread and was savaged by other freepers for not doing my homework.
Personally I think its real and accurate but its not acknowledged by statisticians.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.