The trouble with polls

This morning, while perusing on-line news sites over my first cup of coffee, I was struck by one report on the Washington Post site that declared, "A CNN/Opinion Research poll shows 48 percent approve of Obama when it comes to Ukraine, while 43 percent disapprove."  Gee, that makes it sound like informed Americans, with international diplomacy, government, and security credentials/experience, have thoughtfully and carefully analyzed the the manner in which POTUS is addressing the situation in Ukraine.  However, that is far from the case and illustrates the trouble with polls.

(Disclaimer: the above is just one poll example and is not meant to focus attention on POTUS and the current situation.  Aside from this point, I don't think all of what POTUS is or isn't doing are know to the public.)


The trouble with polls of this nature, wherein a "representative sampling" of folks are asked what is in essence a technical question, is that the vast/overwhelming majority of respondents are unqualified to provide an informed, let alone expert opinion.  Further, as readers of the poll results,we have no idea who was polled and why.  In the above, were those polled from retirement homes in Florida, suburban residents from California, farmers from Iowa, or only those who answered those annoying telephone calls made during the dinner hour?  What makes these Americans, our fellow citizens (assuming the respondants were indeed citizens) qualified to offer expert opinions? 

Nothing.

All that are offered are the option to select from certain responses to precrafted questions meant to elicit certain replies.  And there are no variations in replies allowed.  It's a play on the old trick question, in which you are asked, "You can only answer the following question Yes or No.  Have you stopped beating small puppies?"  Remember, the way the "poll" was asked, you can answer only YES (you have stopped beating small puppies) or NO (you haven't stopped beating small puppies).

So, depending on the way the "poll" questions and available responses are crafted, the results are predestined.  Second, we have no way of knowing the expertise of the respondents.  Third, we have no way of knowing who were polled.

Thus, the process is flawed, particularly when seeking to inform the public about rather important topics.  The above example wasn't like polling to ascertain the favorite flavor of ice cream (chocolate, by the way).  It addressed a rather serious matter.

The trouble with polls of this nature is they don't produce anything of value.