Well, saudade doesn't want to be quoted anymore and have to see rebuttals, so I guess I won't. I wrote a few research papers last semester and the one before that on pitbulls, media and bite and attack statistics, though, as well as breed specific legislation. Obviously I'm pretty passionate about this, even though I don't own a "bad" breed, so I can't not respond.
The problem with things like the CDC study is that there is no "real" pitbull. What is a pitbull? Do tell. We'd all kind of like to know. Is it only an American Pit Bull Terrier? What about all of the mixes, the "bully" types, staffordshires, bull terriers? Does a short-haired dog with a big jaw become a pitbull just because it attacks someone? It really seems to be the case. There have been countless cases of misidentification of breed when people are attacked, simply because people can't tell a pitbull from a cane corso from a dogo argentino from a labrador mix. I saw one article where a dog with one blue eye, relatively narrow jaw and GSD type coloring was being called a "pitbull." I saw another where it looked more like a rhodesian ridgeback cross. Who do you think is deciding what the breed of the dogs are after attacks? Dog experts, long-time APBT breeders? Are you really going to assume that legislators, attack victims and statistic writers are all dog experts? Animals are my life, and even I can't always tell an APBT from a lookalike breed or mix. Do you have any idea how many of these stories report "huge" pitbulls, when APBTs are really about medium sized dogs?
Also, as someone thinking about entering the field of science, I'm somewhat annoyed people cite such a thing as a "scientific study." There are no breed-specific truly scientific attack/bite statistics, exactly because of the aforementioned problems. "It looked like a ___ so it is a ___" has never been a part of any kind of scientific reporting. Unless they personally DNA tested each and every dog to make up the statistics, it's no more scientific than a widely held opinion.
Pitbulls are not bred for human aggression. They are bred for dog-on-dog aggression, which is a very different drive. I've actually been looking for a personal protection dog in the past year, and I know to steer clear from pitbulls- as do all other dog people- because they're NOT human aggressive or territorial enough to do the job reliably. Misguided idiots with an ego problem to sort out think that they'll make good protection dogs because they're "vicious," and then the poor dog ends up in a shelter after years of beating and neglect failed to make it vicious. A fighting dog can't be aggressive towards people, because people are the ones that are grabbing them and dragging them out of the pit after they've been ripping at another dog. If it turns around and attacks its handler, what do you think that handler's going to do? It's not like people who fight dogs love them, after all.
If you're adopting, you have even less to worry about than usual. I've worked with dogs in more than one rescue, and because of the pitbull's reputation and ovepopulation, only the best, most well-behaved ones are put up for adoption. There's just too many of them for them to put up any more than that.
I had a friend who had a hound/pitbull mix she adopted, and the rescue put on his paperwork that he was "hound/terrier" mix. Because he's leggy and brindle, she just tells everyone he's part plott hound, and because most people aren't exactly dog savvy, no one can tell the difference. So if you want a pitbullish type, I suggest getting one that the shelter marks as a "terrier mix," first off, and then making sure it resembles a breed with a more benign reputation, like a lab or something. Boxer/terrier mix or lab/terrier mix is common code for pitbull-type, by the by.