Poll sampling has a liberal bias!

Graph of Bias.

The above graph does not have liberal bias. Quite the opposite: it suggests a brilliant way to fix the liberal bias that can usually be found everywhere in facts, statistics, and data in general.

This graph, produced by a very credible anonymous person on the internet who uses the nickname “NumbersCruncher” who posted it on Twitter and then had it later published on the extremely reliable news source The Blaze, clearly shows that most of the election polls suffer from a “sample advantage” of Democrats over Republicans.

What does “sample advantage” mean? It means that they polled more Democrats than Republicans.  As you can see, every poll except Rasmussen sampled more than 4% more Democrats than Republicans, and every poll except Rasmussen shows Obama having an advantage. The conclusion is obvious: the only reason these polls show and advantage for Obama is that they surveyed more Democrats than Republicans!

Now, the people who do the polls might say something like this: Consistently, likely voters have actually consisted of 3% more self-described Democrats than Republicans, and in the last presidential election the voting population actually was made up of 7% more Democrats than Republicans. Therefore, our survey sample simply reflects the statistical reality of the make-up of the voting population.

But that’s stupid.

Obviously, the only way to have a fair and unbiased survey is to poll half of the people on one side, and half of the people on the other. I mean, that’s just common sense. When there are two sides, the fair thing to do is to survey half from one group and half from the other. Fair means both sides are equal. Nothing could be more obvious. Check your gut feelings: doesn’t that seem fair?

Therefore, we propose that the genius method suggested by this anonymous internet person in the above graph actually be applied even more widely.  For example:

To determine whether global warming is true, survey 100 climate change supporters and 100 climate change deniers.

To determine whether evolution is true, survey 100 evolutionists and 100 creationists.

To determine whether English should be the national language, survey 100 people who believe that English should be the national language and 100 people who think it should not.

Quite plainly, this is the best way to get good, conservative statistical results.

Anything else would just be liberal bias!!!!


Graph Data Source: who knows? some anonymous person made it and posted it on Twitter. That’s how we know it’s true.
Graph Found Via: TheBlaze.com