What a lot of people don't know, is that Nielsen only tracks around 5,000 households to see which shows people watch. Based on the number of households which have televisions (estimated at 99 million), this means that only .005% of television viewers are tracked. Nielsen does their best to make sure they have as good of a sample as possible as to different demographics, but .005% in my world, isn't statistically viable. Those 5,000 households don't hold the answers to which shows are truly most popular.
The reason I mention the above is that we sometimes rely on various sources without knowing how legitimate the information is that we're ultimately believing in. Various news agencies used to have an ethical code they lived by where if they provided information about surveys, that at the bottom of the screen they'd tell you how many people were surveyed. Now they don't. Why?
In most cases survey companies only survey around 3,000 people for each topic. Let that sink in. Why in the world should we believe the media when they say things like: "Seventy percent of Americans believe semi-automatic weapons should be outlawed,"? Maybe it's true, but we certainly wouldn't know for sure based on how many people were asked the question.
They typically only interview .0009% of Americans in surveys like this, yet they act like what they're sharing is the truth. What makes it even worse is they don't tell us "where," the surveys were taken. If you want to prove Americans support investing in green technology, you'd want to survey people from places like California and New York. If you want to show that semi-automatic weapons shouldn't be outlawed, you'd want to conduct surveys in the southeast. I used to work in research and quite frankly, I could find any story you wanted. I could prove that global warming exists and I could prove that it doesn't.
The bottom line is that we're better off going with our common sense and logic regarding various topics than believing the mainstream media about various surveys.