This survey is chock full of bias, including several systemic bias errors. One of the most glaring errors is they ask whether people are in favor or opposed. People don't like to be seen as opposing anything. Thus, the "oppose" category is artificially deflated and the "in favor" category is artificially inflated. On a broader note, both Type I and Type II errors are created because of this and other biases in the overall design as well as the nature of the questions asked.
This brings us to those questions, all of which are given a heavily positive slant. Who wouldn't be in favor of "raising the national wage to $10.10 an hour?" The term "raising" is positive. The hidden negative is keeping people at a lower wage. Another camouflaged, though not entirely hidden positive is the use of the term "national," which has a positive relational index. People favor nationalism as much as they favor the Fourth of July.
Minimizing bias from surveys is difficult, but by no means impossible. Those who specialize in certain sub-fields of statistics, logic, and psychology are well-trained to spot and avoid bias. The vast majority of researchers who do these polls, however, are fairly clueless when it comes to eliminating bias. All to often, they unintentionally (and sometimes intentionally) induce bias, largely as a result of their own preconceived notions or desired outcome.
A proper approach to minimizing bias with respect to the minimum wage question begins with the science of statistics and understanding known biases, but it involves a bit of word art, as well:
The current federal minimum wage is $7.25 per hour. Some people propose keeping it as it is. Some people propose increasing it to $10.10 per hour. Discussions have revealed both good and bad consequences for both options, not only for individuals, but the businesses, as well as our nation's economy as a whole. What do you think is the best course of action overall? A) Keeping minimum wage at $7.25 per hour; B) Increasing minimum wage to $10.10 per hour.
Using the term "federal" is more accurate, as it's the federal government, not a state government, who is making this proposal. Furthermore, although the term "federal" has its own positive connotation for some, it has negative connotation for others. Overall, the net connotation is minimal, as opposed to the strongly positive connotation with the term "national."
By using the term "think" vs "prefer," you're making it less personal. Obviously, individuals would prefer making more money! However, when asked what they think is best, they'll often come up with a more objective conclusion.
By including a brief and neutral setting, that "discussions have revealed both good and bad consequences for both options, not only for individuals, but the businesses, as well as our nation's economy as a whole," you're actually introducing one bias, but countering another bias. The bias you're countering is personal bias. You're minimizing it by changing people's frames of mind from the personal to the objective. The result is a more accurate question with respect to what should actually be done about the issue, as opposed to what people think is best for them.
The two answers both include terms ("keeping" and "increasing") which are equally positive and therefore negating. The result is a neutral option of two answers. This example includes positional bias. To minimize positional bias, half the questionnaires would have "Keeping minimum wage at $7.25 per hour" as answer A, and the other half would have it as answer B.
Finally, it's important to avoid a third option, that "Some people propose raising it to somewhere between," as that also induces a bias, where people tend to gravitate towards the middle, along the lines of the old adage, "the truth is somewhere in the middle." In reality, it's often not. Yet if you include this as an option, it will significantly detracts from both the "keep" and "raise" options, leading to the erroneous conclusion that most people prefer at least some increase, when the actual results may very well be 60% oppose, 30% approve, 10% undecided. That's a very clearly "opposed" response, but if you include the middle option, you may very well bleed half of the answers towards the middle with a 30% oppose, 15% favor, 50% somewhere between, and 5% undecided. Not good! Nor is that accurate. It's biased.
The remainder of the survey is just as heavily biased in one way or another. The very fact that its "researches" chose such an amateurish "in favor" / "oppose" response duality is clear indication these folks haven't a clue as to what they're doing. I don't care how many degrees they have among them. This is NOT proper research. This survey is heavily flawed, and not to be trusted.
Because it is seriously flawed, it most certainly does not show support for gun control, much less "massive" support for gun control. All it shows is that the folks who designed, administered, and analyzed the survey and its results were complete blithering idiots when it comes to the proper use of statistics.