• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

Are online polls legitimate?

Hercules56

Banned
H
Joined
Aug 4, 2013
Messages
17,176
There have been several online polls about the election

Many of them say that Trump is beating Hillary by HUGE numbers.

One poll gave Hillary only 10%.

Are these polls legit at all?

How do we know which online polls are legit?

In my view the only online polls that are legit are ones that block proxy servers and hide cookies in a hidden place on one's computer so they cannot be easily erased.
 
 
Legit for what?

They're a great way to measure what the participants think, as the backend work can be easily automated, and results tallied and displayed according to the user's desires. They're worthless when it comes to gauging any other population, however.

For example, poll results for a thread on this board can give you a reasonable idea of what the thread's participants are thinking. But if you stretch it to "all Americans" or "All janitors in Germany" or "all dogs", the poll becomes invalid.
 
Mumbles said:
Legit for what?

They're a great way to measure what the participants think, as the backend work can be easily automated, and results tallied and displayed according to the user's desires. They're worthless when it comes to gauging any other population, however.

For example, poll results for a thread on this board can give you a reasonable idea of what the thread's participants are thinking. But if you stretch it to "all Americans" or "All janitors in Germany" or "all dogs", the poll becomes invalid.
Click to expand...

Yeah, I pretty much agree with this. Well-spoken sir. +15 internet points for you.
 
The voluntary response polls partially reflect the intensity of support Trump's receiving. Someone on a blog I follow said that Clinton did a better job in the debate but he did his "part" by voting for Daddy Trump in five different polls.
 
Brainster said:
Remember Ron Paul winning all the online polls in 2008?
Click to expand...

No, but I have no doubt it happened.

Bernie's boys posted messages all over Facebook, telling folks to flood the CNN polls to make Bernie win.

Which of course, shows how these polls are actually craptastic.
 
The web ate my post, I shall try again.

Online polls aren't worth the paper they're not printed on. Safeguards against voting multiple times can easily be defeated by deleting cookies or using incognito windows.

In another thread, it was mentioned that P.Z. Myers told his supporters to go flood a creationist poll. His reason for doing that was actually to show how worthless such polls are. The idiots running the poll kept restarting it because it didn't give the results they wanted.
 
Mumbles said:
Legit for what?

They're a great way to measure what the participants think, as the backend work can be easily automated, and results tallied and displayed according to the user's desires. They're worthless when it comes to gauging any other population, however.

For example, poll results for a thread on this board can give you a reasonable idea of what the thread's participants are thinking. But if you stretch it to "all Americans" or "All janitors in Germany" or "all dogs", the poll becomes invalid.
Click to expand...

Not even that. A poll result on this board might give you a half decent idea of what the thread participants think because (and I could be wrong) you have to register to vote, and this is a fairly obscure forum.

But a post on a news site is high profile and publicly accessible, which means it's not what the news site's visitors think, it's what the chan/reddit board mob and their bot-programs think.
 
Trebuchet said:
The web ate my post, I shall try again.

Online polls aren't worth the paper they're not printed on. Safeguards against voting multiple times can easily be defeated by deleting cookies or using incognito windows.

In another thread, it was mentioned that P.Z. Myers told his supporters to go flood a creationist poll. His reason for doing that was actually to show how worthless such polls are. The idiots running the poll kept restarting it because it didn't give the results they wanted.
Click to expand...

If this is the one I recall: they kept restarting because the volume of traffic kept crashing their server. The poll didn't even have the most basic anti-spam configuration. Theoretically, you could just press the back button on the browser and then the vote button to vote again. I was part of one team that worked out an optimized auto-vote program (minimum packet lengths, alternating intermediate routers, no-return packet, etc). I know that there were at least three such teams, we started competing with each other to get the highest votes/hour count. I'm sure redit still has the thread somewhere...
 
On line polls are for entertainment value only. Anybody who took Statistics 101 (a required course in most colleges for a Business Adminstration degree) can tell you they have no controls, no scientific sampling, and are useless.
 
I believe there is such a thing as a scientific online poll, but not the ones where the respondents self-select.

It's getting harder these days for traditional polling companies like Gallup to get a truly representative sample by calling people on landline telephones. Some younger people these days don't even have such. They have a mobile phone so they figure who needs a landline. So polling companies are looking for new methods to keep up with the times.
 
The only people who "believe" in on line polling are the ones who win them at that moment. The above link to Jason Miller citing them is a perfect example.

Hell, we've had threads here where WE, as in "members of the forum" manipulated on line polls. Rebecca Watson won some popularity contest to try to gain a guest commentator slot, we won something for Lisa (yes, I'm blaming Lisa) for her cat, we fought to the wire to get Phil Plait (at his request) up-voted for Science Blog of the Year (which he lost to Peezers) and we fought back in numerous b.s. truther polls. Included in all those threads was an instruction as to how to manipulate it if we could find it.

They're nonsense and easily manipulated. As I mentioned in another thread re the election night polls, Trump's bots/supporters had the past tense "Who won" vote in there during the opening twenty minutes of the debate, as soon as the polls opened.

The only way to view on line polls is as a sampling of how enthusiastic some people are about winning polls. It could be for the lulz it could be for gain. Sometimes the public just gets behind a funny one (Boaty McBoatface comes to mind*) but generally it's a concerted effort by someone, with the "one" in that representing groups.

*When Charles and Diana had their first son, the heir, a television station in Cincinnati ran a call-in poll. The winner? "HRH Prince Bubba".
 
Last edited:
Hercules56 said:
There have been several online polls about the election

Many of them say that Trump is beating Hillary by HUGE numbers.

One poll gave Hillary only 10%.

Are these polls legit at all?

How do we know which online polls are legit?

In my view the only online polls that are legit are ones that block proxy servers and hide cookies in a hidden place on one's computer so they cannot be easily erased.
Click to expand...

The problem is one of how the sample set is determined. Just having one open on the interent opens up to all kinds of issues, like the Colbert module on the space station and Boaty McBoatface for research vessel. So you need an effective way of getting a sample that is actually representative. And with one just open to the public you are not, you are going to get some groups vastly over and under represented.

Now if you sent the link out to say everyone in a certain organization and limited it to that it would be a great method of polling. It is all about how you limit or do not limit sampling biases.
 
You must log in or register to reply here.

Back
Top Bottom