Models —- How To Skew The Polls


By now, we’ve all seen and heard the polls telling us this election is all but over.  And, it is true, most of the polls are saying that Obama is opening up a lead.  Even the polls I trust are saying this.  Pew and Rasmussen show Obama in front. 


Although when you look at it, this shows Obama at just over 48% and Romney at 46%.  This is within the margin of error, so the race is essentially tied.

Most people reading this blog are aware of models in a different context.  We’ve seen models used to predict our climate.  And, they’re always wrong.  This doesn’t necessarily mean the people running the models are manipulating them to get the outcome they want, although, we have seen some that have.  This simply means the models are weighted wrong. 

In regards to the climate, most models give inordinate weight to CO2.  It screws up the results because the inherent weighting errors of the model. 

So what does this have to do with the polls?  All polls are filtered through models.  And, we see some obvious flaws in some of the more slanted polls, such as predicting a heavier Dem turnout in 2012 than 2008.  We know this isn’t going to happen.  The polls which do so can be automatically ignored.  Quinnipiac and Marist do this so don’t worry about what they have to say.  But, what about Pew?

Pew is trickier.  Pew is typically pretty spot on.  On the 27th, they released a poll which CNN covered. 

Pew poll: Obama opens up lead over Romney among Catholics

Now, this caught my eye, because I’m incredulous about this.  I’m aware that many Catholics are traditionally Democrats, but the callous actions of this administration towards Catholics should move them over to Romney.  Further, this poll reinforced what many of the other polls are saying.  It has Obama over Romney 51% to  43% among likely voters.  Wow!  So what gives?

Let’s look at who they sampled. 

GENDER registered voters/likely voters
Men              1097  /  984
Women        1327 /  1208   That’s 55.1% women.

Now, we know this is wrong.  Women had 53% of the vote in 2008.  But, there’s something even more wrong with this poll.

Republican  717 / 665
Democrat 869 / 803
Independent 757 / 661

I haven’t run down how Pew filters this through their models, and it doesn’t matter.  This is the other side of the equation.  GIGO is what this is.  Somehow, Pew was able to dial the phone and get more registered voters who are independent than Republican.  Really?  Not a chance in hell. 

So, yes, when asking nearly 70% of independent and Democrat registered voters, Obama may be gaining in the Catholic vote, and Romney is probably behind 51% to 43%.  Add a little gender skewing and voila!  Romney is getting beat.  I want to be clear, in this particular instance, I’m not suggesting malfeasance, I’m suggesting incompetence.  The sampling is wrong so regardless of the model, the output will be wrong. 

Some notes about some of the polling. 

When “leaners” are included, it’s Obama 49%, Romney 47%. Leaners are those who are initially uncommitted to the two leading candidates but lean towards one of them when asked a follow-up question. Beginning this Monday, October 1, Rasmussen Reports will be basing its daily updates solely upon the results including leaners.

I can’t find the quote, but I believe Gallup will make a similar change as well. 

You can read the Pew poll here.  (pdf)

This entry was posted in News and politics. Bookmark the permalink.

10 Responses to Models —- How To Skew The Polls

    • suyts says:

      I just can’t believe Pew went with those numbers. It’s very out of character.

      • kim2ooo says:

        Actually, no.

        I have watched the Pew Religious Polls for awhile now. Many of us have questioned their
        figures for a number of years. Especially pertaining to Catholics.

      • suyts says:

        Well, I must confess to not following their polling of religious people. I was stating that they generally are pretty close on their polling in the general elections.

        • kim2ooo says:

          I dono if one can find the Pew Religious Polls pertaining to Catholicism and the Obamacare anymore.

          They were a big mover of the Stupik 12 ( Democrat Catholics )

        • suyts says:

          I think I’ll keep a close eye on this. I’m very skeptical that Catholics are moving towards Obama.

        • kim2ooo says:

          suyts says:
          September 29, 2012 at 3:59 pm

          I think I’ll keep a close eye on this. I’m very skeptical that Catholics are moving towards Obama.

          I think you are right! 🙂

          If you send me an email – I’ll give you a link

  1. Pingback: Oops! Champaign Popped Too Quick—- More Polls | suyts space

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your 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 )

Connecting to %s