Search This Blog

Wednesday, March 8, 2017

It Really Just Cannot Be Correct

Today brought a series of polls on President Trump's job approval by the American people.  As has been the case since Trump took office, the poll numbers do not add up.  For example, the Quinnipiac results show an 11%  margin for disapproval of Trump while both USA Today/Suffolk and Rasmussen show 3-4% margin for approval.  That means that one or the other is wrong.  Since the start of Trump's term, Quinnipiac has always shown disapproval by a large margin.  The last Quinnipiac poll before this one showed a 17% margin for disapproval while the data of the other pollsters was much stronger for Trump.  I've checked, and the question asked by Quinnipiac is not slanted compared to the other pollsters.  There are only two ways that the Quinnipiac numbers could be so different:  either the sample used is biased against Trump or Quinnipiac is just making up the numbers.  I assume that the cause is the sample.

Pollsters usually adjust the results they get from polling to a proper sample.  For example, if the people questioned are 35% GOP voters and 41% Democrats but the pollster thinks that the electorate is actually 33% GOP and 39% Democrat, the results are adjusted to fit the expected electorate.  There has to be something in the adjustments that Quinnipiac is making that is skewing its results so far from the those of the other pollsters.

This may sound like a rather technical point of little importance.  It isn't.  Every time that Quinnipiac does a poll that shows low numbers for Trump, the media tells America that Trump has the lowest poll numbers of any president at this point in his term.  The polls that show Trump much more popular are rarely reported.  It's important to understand that the polls on which the media focuses are bogus. 

No comments: