Figures Don’t Lie, But the “Figurers” Do!

 

Over the past few months, we have seen a variety of polls – some with significant swings regarding who is in the lead for President. The lame stream media never fully explain the impact of how the polls were handled, whether there was over-sampling, whether registered or likely voters were surveyed, or whether the poll was conducted over one day or several days. All of these factors, and more,  make a difference in whether the poll is an accurate snapshot of the population as a whole. The figures do not necessary lie, assuming they are reported accurately from the interviews conducted. However, the “figurers” – those creating and reporting on the polls – do.

Polls

Photo credit: methodshop.com

For instance, today, Gallup released a poll with the title, “Voters’ Reaction to Romney’s ‘47%’ Comments Tilts Negative — Independent voters, by 29% to 15%, also more negative than positive.” At first glance, it sounds like this is a horrific turns of events for Mitt Romney — all is lost, the independents have turned! But wait, not so fast.

Gallup explains on their own website that “not all of these registered voters will actually vote.” In addition, Gallup claims to have “created systems to isolate likely voters.” Why, then, are they running a poll using all registered voters, knowing fully that this is not the best representation of the actual voting pool? This is just one method to acquire the poll headlines that best suit the organization’s interests (especially when you’re threatened by the Obama DOJ) — ignore turn out, or use out-of-date turnout models, and base your poll on everyone registered to vote irrespective of their actual voting history. If someone is registered and rarely votes, their opinion is not nearly as significant as someone who is likely to vote, such as an individual who has voted in two or three of the last four elections in their area.

In this specific Gallup poll, 15% of independents state they are now more likely to vote for Romney because of his comments, and 53% stated it did not make a difference at all. In total, 68% of independents said they either did not care or are now more likely to vote for Romney. However, the headline only focuses on the 29% who said they are less likely to vote for Romney, forgetting that 68% either lean toward Romney or stated his comments do not change their opinion (whether for or against).

Yesterday, there was a Washington Post poll showing Obama leading Romney by 8 points in Virginia. Again, the figures themselves were likely accurate, but hark! – the “figurers” were certainly less than truthful by using a sample that favored Democrats by 8 points (32% D, 35% I, 24% R). By way of comparison, the 2008 elections that brought out all of the Obama voters show that Democrat turnout was only 6 points higher than Republicans. In order for the Washington Post’s model to be representative of the actual political environment, we would have to assume that even more people are going to turn out for Obama in Virginia this November, when the enthusiasm gap favors the Republicans.

While these are only a few examples, these lend themselves to the saying that “figures don’t lie, but liars can figure.” Organizations interested in showing Obama in the lead are using a variety of tools and tricks to create the results they desire. Consumers of media must ask themselves who was polled (registered voters or likely voters), whether the sample reflects the anticipated turnout (the percentage of Republicans versus Independents versus Democrats) and whether the poll was conducted on one day or over a few days. Unfortunately, you will often find answers that are disheartening.

 

Find organizations that consistently use a methodology providing an accurate snapshot of the population, and disregard the other outlets. While the lame stream media may continue to report the statistically irrelevant polls, you can be an educated consumer of the media and not allow yourself to be swayed.

 

Trackbacks

  1. […] takedown of the proliferation of skewed polls: Figures Don’t Lie, But the “Figurers” Do! Share this:TwitterFacebookLike this:LikeBe the first to like this. This entry was posted in […]

Leave a comment or question. Report abusive, harassing or annoying behavior by clicking on Tips/Contact in the top menu.

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: