Polling is an extremely difficult thing to do in any sort of objective manner and incredibly easy to do in a way that ensures getting the answer the poller wants. That’s why advocates of all stripes love using polls to “prove” the public backs their position.
First, most people won’t bother to respond to a poll which is too long, too detailed in its questions, or requires much more than yes or no answers. Once the questions and possible answers have been dumbed down enough to be attractive, or at least not too annoying, to the intended audience, then the meanings have been distorted or made so ambiguous the the results are worthless in any scientific sense.
Second, even with the most rigorously constructed poll it matters who is asked to respond. Pick the right persons to take the poll and you’re bound to get the answers you want.
Third is language. The words used in the questions, their connotations, the emphasis placed on certain words, the overall implied tone, are all tolls to elicit the desired result.
Consider these versions of the same question:
Do you believe the 2nd Amendment guarantees the right for a law abiding citizen of legal age to possess and carry a firearm for self defense?
Do you believe the 2nd Amendment grants the right for anyone to buy a gun and carry it in public, provided they aren’t a felon?
Do you agree that the 2nd amendment was intended to allow ordinary, unregulated people to own and carry guns whenever they want to?
Do you believe the 2nd amendment gives people the right to bring deadly weapons into public places such as concerts, shopping centers, parks, flea markets and sporting events?
Despite the 2nd Amendment, do you believe people should be allowed to have military-style weapons in their homes, their cars, or hidden under their clothes?
Now, you and I might notice the flaw in some of these questions, but we have spent some time and effort to research and understand the meaning of the 2nd, as well as what it meant to the people who wrote it. The general public is often not so well versed in the nuances of “guaranteeing a right” vs “granting a right”. Aside from that, the questions are broadly accurate but run the gamut from fairly objective to aggressively biased.
Now consider where the poll is presented. By someone knocking on your front door? By someone on a street corner? At a rally or event intended to promote one side of the argument? Through the mail, snail or electronic?
How about who is offering the poll. A well-dressed young adult, polite and smiling? An activist displaying their point of view with graphic clothing, signs, banners? A bored, minimum wage temp worker sitting at a card table at the entrance to a supermarket?
How about when? The day after a horrific mass shooting has occurred and received repetitive, round the clock news coverage? After one side experiences a major win or loss in the courts? During a time of relatively low reports of incidents? On the heels of a well publicized “good guy with a gun” success story? In public or in private?
Finally, and most importantly, WHO is asked to respond to the poll. A random sampling of people on the street? Which street, which city, which state? At a location where the majority of people present are likely to be biased one way or the other? Registered voters? Academics? Students? People in high crime areas? Low crime areas? Economic status? Education level? Occupation? Age? Religion, either by denomination or lack thereof? The list goes on and on.
The real question to be asked of any poll/methodology is this: Is whoever sponsored the poll trying to learn something or prove something? It all starts there.