2016 Postmortem
Related: About this forumPersonally I think someone should be fired for that bogus poll from NYTimes\CBS
I'm not a professional poller but I have taken a class that deals with how to gather fair samplings to ensure viability of your numbers. What NYT\CBS did with this recent poll 'claiming' that Mitt Romney somehow is leading with women was nothing more than pure bullshit and someone should be held accountable for it.
In a nutshell NYT/CBS decided to use the same people they polled about a month ago to complete their survey. The original survey was with 900 people, a pretty good random population. But for the most recent one they only used the same 900 people again for the survey. That wouldn't be an issue except they only were able to get about 60% of them to respond back this round (around 575). That's a 40% loss from the prior polling which means your data is extremely flawed. The only way to fairly do this poll was either keep doing the poll until you reach all 900 people in the original one or just do a sampling from the same master list of people.
And maybe it's me but I've always seen the New York Times and CBS as 2 fair organizations but they really blew it with this poll. This poll only does one thing - creates headlines that says 'oh the race is close'. Let's face it a close race means 2 things - better ratings for the nightly news and the better chance of having a stolen election.
Shame on you NYTimes and CBS - shame shame!
Dawson Leery
(19,348 posts)This is the way Fox "News" Channel skews their data to favor their party, the GOP.
get the red out
(13,468 posts)Gotta keep the horse race going, even if it's a mule alongside a Kentucky Derby winner.
Wilms
(26,795 posts)You nailed it.
liberal N proud
(60,351 posts)One that will take place in November, like say on the 6th of November.
You have to have polls that can be used to convince the public that the election is not stolen. So leading up to the election they media and pollsters all have to do whatever they can to support the pending crime.
HowHasItComeToThis
(3,566 posts)LoisB
(7,249 posts)If the media can convince us that it's 47-46 or 48-48, we'll be less inclined to question a Romney win.
annabanana
(52,791 posts)Except this time we're on to them.
sharp_stick
(14,400 posts)justification for doing a poll like this. My guess is either they wanted to save some money or they really wanted to skew the results and this is how they could do it.
If it's the first, they shouldn't bother doing the poll until they've improved the budget situation.
If it's the second at least a couple of people should be out of a job, maybe that would fix their polling budget problem.
Arkana
(24,347 posts)The percent of Republicans sampled goes up 5%, the number of Democrats up 2%, and the number of Indies drops 7%.
HowHasItComeToThis
(3,566 posts)GOTTA MAKE A BUCK.....
patrice
(47,992 posts)in public education.
SoutherDem
(2,307 posts)I can only see using the same sample twice to show change in opinion as in after new information is given. In such a process you realize you may not get the full sample again, but take that into consideration when determining the results. To use the same sample is simply lazy and does give the impression of not being ethical. Once you have the feeling of a particular sample it could easily be said that you knew the direction that sample and would affect your second poll.
I am concerned, as you are, that by using the original sample and relying on only those who could be reached during the time of the polling does in no way give accurate result. There may be a reason to sample a sample, but I don't see where this would apply. Also, if you are going to use this process you need to make sure it the distribution of the small sample is the same as the full sample, if you are going to claim it still represents the opinion of the full population.
I however am not concerned with the smaller sample 575 vs. 900. 575 if well distributed to match the target population would/should give the same results of the target population.
That said pollsters are not always the most ethical people. It is very easy to direct the results by the way the questions are worded or even the time of day you are conducting the poll. In both my basic statistics class which used a business statistics book and advanced statistics class which used a psychology statistics book looked into ethics. I would have never been able to use this process for even a project one of those two classes much less a national poll use by two major media groups.
I feel many polls are for driving the headlines or in FOX News case support their position.
Jennicut
(25,415 posts)Isn't that biased polling? It is not a clear, well rounded sample. It seems like it was done on purpose to get the results they wanted.
SoutherDem
(2,307 posts)Last edited Tue May 15, 2012, 04:43 PM - Edit history (1)
and may have been on purpose.
I am not saying is isn't. It very well may have been. I just won't automatically say that due to the reasons listed, reusing a sample and accepting only a 60% reply rate.
I haven't seen the full results (only the summary), if it is even available, so I won't say that for sure if it was biased, lazy or actually used acceptable methods.
If it was a good sample that really represented the population, to reuse it may not be that bad, although knowing how a sample leans can generate desired results. To set a time limit and accept a 60% reply rate at the end may not be bad if a good representation of the total population replied and to extend it could also point to bias.
However, unless they closed the poll once they had the results they wanted, I don't feel reusing a sample or accepting only 60% reply rate as proof of bias.
There are a lot of easier ways to get desired results. By the way questions are worded, dialing only certain exchanges, selecting the times of day you call are all ways of getting desired results.
When it comes to polls we all, myself included, will accept polls which agree with our opinions and dismiss those which don't. I simply don't have enough proof to say it was biased, except to say I don't like the results.
Response to LynneSin (Original post)
Post removed
Bigredhunk
(1,351 posts)N/M
LynneSin
(95,337 posts)even though you barely knew the material?
That's what the poll is - a bunch of bullshit, poor sampling created to help keep the race close. You believe stuff like this poll then on election day you're gonna wonder what happened - how did he lose?
I'd rather know the truth with fair and correct sampling method - but hell, if the poll makes you feel good then so be it.
MiddleFingerMom
(25,163 posts).
.
.
I haven't gotten out-of-hand with MY polling.
.
.
.
In FAR too long, dang it!!!
.
.
.
.
.
.
Iliyah
(25,111 posts)I mean ti seems like there are hundreds, and seems most are RW ones. I'm tried of them.
People get out and vote for Pres O and other dems!
ailsagirl
(22,904 posts)Even if they have Obama ahead-- I-just-don't-trust-them.
marlakay
(11,527 posts)they are going from coast to coast mid country to move and she asked me today is it true Romney is leading in the polls? that is what i am hearing on the radio today.
She was in mid Missouri headed to kansas when we talked.
I laughed and said where you are I am not surprised! she said its all country music too
luckily she likes that.
I felt better when she said, I thought about voting for Romney since my husband is but I just can't do it..I am voting for Obama. My son n law started out a dem but after ten years in military he has slowly evolved
the wrong way!!
I didn't notice a big change though until his last 3 years at the pentagon
that is what swayed him
TwilightGardener
(46,416 posts)from it, and the weak-minded will eventually be persuaded.
Forward2012
(14 posts)in my experience. Of course I was around all the scientist-types and they have some pretty far out theories on how governments should run.
marlakay
(11,527 posts)When it got to the point that all the people he worked with, was friends with we're republicans it slowly wore on him. And he is real into money, dreams of having money, always been that way.
coalition_unwilling
(14,180 posts)about WMD in Iraq uncritically and sycophantically. As such, neither retains any credibility whatsoever for any story of substance. Really a shame, as Seymour Hersh broke the My Lai massacre story in the NY Times in a different era. And Dan Rather used to report critically from Vietnam as a reporter for CBS during the Vietnam War. Oh, wait, CBS fired Dan Rather for completely bullshit reasons.
A lot of people died, were wounded or truamatized because of their shoddy journalism.
Fuck CBS and the NYTimes. They're both a disgrace to anything resembling serious journalism.