Tl;dr: The dogsbite data set is poorly constructed and misleading. Real science performed by actual scientists has overwhelmingly found the belief in inherently violent bully type breeds to be exactly that - just a belief.
This whole thread is a great example of why science education is so important. I suspect a not insubstantial percentage of the people touting the “70% of dog bites are by pitbulls” statistic have previously mentioned how correlation is not causation and decried the cherry picking of data.
Here we have many individuals cherry picking data while making claims that require correlation to be causation. They also seem to lack the skills required to evaluate the strength of a study or data set, otherwise they wouldn’t be quoting one where calling it shoddy would be a massive compliment. My cat makes better, then scratches dirt over it.
For anyone interested in actually learning about the nuance of the situation, plus why the dogsbite data set is pure horse shit, here are a few pertinent articles.
Update: Time Magazine’s article The Problem with People, not Pit Bulls is a great read to get you thinking. Keep in mind the author has a very obvious bias, so keep on your skeptic hat and question all of it.
Uneducated? Yes, at least in math and/or statistics. Stupid? Not necessarily.
My education once stopped at the high school level and I can guarantee you I’d have made the same unsupported conclusions based on the data. I later attended college for nearly a decade and it opened my eyes to a world of bad and misleading statistics. All it took is a little education and a lot of critical thinking.
Tl;dr: The dogsbite data set is poorly constructed and misleading. Real science performed by actual scientists has overwhelmingly found the belief in inherently violent bully type breeds to be exactly that - just a belief.
This whole thread is a great example of why science education is so important. I suspect a not insubstantial percentage of the people touting the “70% of dog bites are by pitbulls” statistic have previously mentioned how correlation is not causation and decried the cherry picking of data.
Here we have many individuals cherry picking data while making claims that require correlation to be causation. They also seem to lack the skills required to evaluate the strength of a study or data set, otherwise they wouldn’t be quoting one where calling it shoddy would be a massive compliment. My cat makes better, then scratches dirt over it.
For anyone interested in actually learning about the nuance of the situation, plus why the dogsbite data set is pure horse shit, here are a few pertinent articles.
American Veterinary Medical Association - The Dangerous Dog Debate
Applied Animal Behaviour Science - Breed differences in canine aggression
American Psychological Association - Pit Bulls and Prejudice
BMC Genomics - Genetic testing of dogs predicts problem behaviors in clinical and nonclinical samples.)
Science - Ancestry-inclusive dog genomics challenges popular breed stereotypes
Update: Time Magazine’s article The Problem with People, not Pit Bulls is a great read to get you thinking. Keep in mind the author has a very obvious bias, so keep on your skeptic hat and question all of it.
If someone actually needs to have these statistical problem with dog bites explained to them, it’s because they’re fucking stupid and uneducated.
Uneducated? Yes, at least in math and/or statistics. Stupid? Not necessarily.
My education once stopped at the high school level and I can guarantee you I’d have made the same unsupported conclusions based on the data. I later attended college for nearly a decade and it opened my eyes to a world of bad and misleading statistics. All it took is a little education and a lot of critical thinking.