The Last Accurate Poll?

A recent study by the Pew Research Center confirmed what professional pollsters have known since 1997:  the traditional political poll, conducted by telephone, is becoming irrelevant.  In fact, the most accurate result to come from telephone surveys, is the statistics on the efficacy of telephone-based polls.  It turns out that less than 9% of people pollsters attempt to contact by phone actually participate in their surveys.

At Keirsey Research we first suspected the accuracy of telephone-based surveys 4 years ago when we noticed the wild fluctuation in the national polls during the 3 months prior to the November 2008 election.  The variance in these polls is now so great that pollsters have taken to running “polls of polls” – that is, they average all the different polls hoping to get to an accurate result.  What they end up with is a “poll” that shows narrower gaps between the candidates than actually exist.  Most pollsters are currently showing the race between President Barack Obama and presumptive Republican candidate Mitt Romney  as statistically tied (National popular vote).  Keirsey Research shows that this is quite inaccurate, and Mitt Romney has major concerns.

At Keirsey Research, we regularly conduct random sampling of people who have completed the Keirsey Temperament Sorter-II as part of our research to determine the relationship between temperament and different choices people make in many areas – including politics.  We recently sampled more then 1600 registered U.S. voters who are planning on voting in the 2012 Presidential election to find the relationship between Temperament and a person’s views on same-sex marriage – a timely topic due to President Obama publically proclaiming his support, and former Governor Romney declaring his opposition.  As part of the survey, we asked people what affect the candidates’ positions on same sex marriage would have on their vote in the 2012 election.  As noted above, the response is not encouraging for Governor Romney.

As we have noted in previous political / temperament research articles, temperament and gender are the most dramatic factors we find in determining political opinions and choices.  The abstract communicators (Idealists and Rationals) tend to be significantly more liberal than the concrete (Guardian and Artisan) communicators.   Just as most public researchers regularly note, women are more liberal than men.  It should come as no surprise then, that if voting were limited to Idealist and Rational women, Barack Obama would win the 2012 election in a landslide, and same-sex  marriage would be legal through the U.S.  However, these groups are balanced by Guardian and Artisan men – while these groups narrowly favor Mitt Romney over Barack Obama, and are fairly evenly split on the issue of same-sex marriage, there are many more of these voters (35-40% of the population) than the abstract-type women (10-15%).

Here are the results of 3 questions in our sampling:

Question:  In the past week, President Obama has for the first time for a U.S. President, stated publically his support for same sex marriage.  Do you agree with President Obama’s position on this issue?

We divided the respondents by temperament and gender, giving us 8 groups.  All groups favor the

Click for full-size graph

President’s position, although less than 50% of Guardian and Artisan men are in support (14% and 19% are undecided, respectively).  Rational women are overwhelmingly in support the right of same-sex couples to marry at 82%, with only 7% opposed.  Overall, it appears that Governor Romney’s decision to soft-peddle his opposition is a smart one – and the follow up questions re-enforce this.

Question: When it comes to deciding whether to vote to re-elect Barack Obama as President of the U.S. in 2012, how important is his position on the issue of same-sex marriage?

The purpose of this question (and the follow on for Governor Romney) was to determine whether the issue made a difference for the candidate beyond his base.  The available answers were:

  • I would vote for him anyway
  • I am more likely to vote for him
  • I would not vote for him anyway
  • I am less likely to vote for him

It is clear that the President’s going public with his support does nothing to hurt him in his re-election bid.  6

Click for full-size graph

of the 8 groups already favor him over Governor Romney, and the undecideds break in his favor in all 8 groups when it comes to the same-sex marriage issue.  By the way – there is not a missing bar in the “NT Women” grouping.  Zero percent of this group that is undecided are less likely to vote for President Obama because of his same-sex marriage stance.

Question: It appears that Mitt Romney is likely to be the Republican opponent of President Obama in the 2012 election.  He has publically stated his opposition to same-sex marriage.   When it comes to voting for Mitt Romney in the 2012 election, how important is his position on this issue?

The converse of the previous question was to measure the impact of Governor Romney’s opposition to

Click for full-size graph

same-sex marriage on his electability beyond his base.  Based on the response to the question regarding President Obama, it would follow that Governor Romney’s stance would have no positive, and perhaps a negative, effect on his support.  In fact, his stance seems to be quite harmful among undecided voters.  What is especially striking about this graph, is that it shows that while more Guardian and Artisan men oppose than support re-electing President Obama, this does not translate to support of Governor Romney.  Romney’s support among these 2 groups is softer than the support for Obama, and it is next to lifeless among all Rationals, Idealists, and Guardian and Artisan women.

In order to correlate our temperament / gender results further, we also broke down our results by age and

Click for full-size graph

gender.  This graph breaks out the support of President Obama, with men and women broken out in 5 age groups:

  • 18 – 24
  • 25 – 34
  • 35 – 44
  • 45 -54
  • 55 – 64

With the results of these questions, the clear strategies of the candidates begin to emerge:  Governor Romney needs to build an election day organization that focuses on getting older males to vote in high numbers.  If he can discourage Idealists and Rationals from going to the polls, all the better for him – but he’ll have to do this through his messaging strategy, convincing the NT’s and NF’s that their votes don’t really matter, so they shouldn’t bother.  For President Obama, his election day organization will be working to get every woman and every person under 35 to the polls.  These are large gaps to exploit, so he’ll most likely be hammering themes from now until November that differentiate him Governor Romney for these voters.

Back to the title of this article.  We believe that our methodology, which randomly samples a U.S. census-balanced (by age, gender, and geography) population of people that complete the Keirsey Temperament Sorter-II, provides more accurate results than any telephone based poll.  Our participation rate among candidates is greater than 60%.  Next week we will publish the first of our regular tracking polls for the 2012 election, so stay tuned.

5 thoughts on “The Last Accurate Poll?”

  1. Youse guys and yer polls are full o’ beans! I’m an INTJ female and NONE of your data represents my feelings, period!

  2. Polls are designed to find averages. Like the Kiersey system for that matter. I am an INTP and it doesn’t represent my feelings either.

    I think one problem with polls and question system is that they cannot get in nuances. For instance the Kiersey system usually has things like “is it more important to be truthful or loyal” or “is it more important to be just or merciful” when the answer in both cases is “both are important”, “it depends” and “you can’t answer that without having room to write a whole essay”. For that reason I considered the descriptions more important in grading myself INTP then the questions.

    The same applies with polls. A lot of people have complicated, nuanced, and sometimes inconsistent opinions about the subject on the question which do not come out easily in polls. Rationals probably have an easier time abstractly framing their views, but their views are still complex even if they know what they are more easily.

  3. As an ENTP, I can’t see the logic of TT’s comments. It may be that her feelings (just using that word in response to poll results makes me think she is a mis-typed INFP) – are represented by a minority response. Since the questions allowed a full range of answers, her response is non-sensical. A logical statement would have been “the majority data for INTJ females does not represent my positions”.

Leave a Reply

Your email address will not be published. Required fields are marked *