"For protection during political activities, such as while voting...." Why did they throw this in?
I can't recall any recent stories about violence or potential violence at polling places that would compel a person to want to carry a gun to the polls. There was an incident of alleged voter intimidation involving the New Black Panther Party in Philadelphia in 2008, but that was quite a while ago.
I wish I knew. They say in the article that they matched their questions to another survey done by UC-Davis. That data is likewise privately held, which is a problem in the gun studies area. Much of the data being collected about gun owners is privately held.
Another part I take issue with is "For protection against police violence." I can't imagine too many people would think this unless they're criminals. Using a gun to "protect" yourself against the police is probably a recipe for becoming a statistic - as a "victim" of gun violence!
Thank you. I feel like I'm back in my philosophy of research or whatever that class was called 🙂 I guess technically there were two of them in grad school. One was just supposed to be research methodology but we spent an awful lot of time talking about the philosophy behind it. Anyway, great post!
It is always an issue in social science research, be it in your field or my own, psychology. We are measuring what amount to latent variables; if we cannot observe behavior to make inferences, then we are left with finding other ways to measure them. It leads to multiple layers of "science". First, we have the phenomenon of interest, then we have to science of psychometrics, of measurement. Some like to pretend that the measurement issues are bias free. But they are not, so they add an additional level of error.
I suspect that the researchers are designing their study and measures in a manner that fits their a priori hypotheses. The interesting thing to consider, which we have found in loking at how measure structure can influence responding (for instance, item order), is how early items in a survery can influence responses to later items. Early items create a context within which later items are interpreted.
"For protection during political activities, such as while voting...." Why did they throw this in?
I can't recall any recent stories about violence or potential violence at polling places that would compel a person to want to carry a gun to the polls. There was an incident of alleged voter intimidation involving the New Black Panther Party in Philadelphia in 2008, but that was quite a while ago.
I wish I knew. They say in the article that they matched their questions to another survey done by UC-Davis. That data is likewise privately held, which is a problem in the gun studies area. Much of the data being collected about gun owners is privately held.
Another part I take issue with is "For protection against police violence." I can't imagine too many people would think this unless they're criminals. Using a gun to "protect" yourself against the police is probably a recipe for becoming a statistic - as a "victim" of gun violence!
Thank you. I feel like I'm back in my philosophy of research or whatever that class was called 🙂 I guess technically there were two of them in grad school. One was just supposed to be research methodology but we spent an awful lot of time talking about the philosophy behind it. Anyway, great post!
Methodology always includes some epistemology!
It is always an issue in social science research, be it in your field or my own, psychology. We are measuring what amount to latent variables; if we cannot observe behavior to make inferences, then we are left with finding other ways to measure them. It leads to multiple layers of "science". First, we have the phenomenon of interest, then we have to science of psychometrics, of measurement. Some like to pretend that the measurement issues are bias free. But they are not, so they add an additional level of error.
I suspect that the researchers are designing their study and measures in a manner that fits their a priori hypotheses. The interesting thing to consider, which we have found in loking at how measure structure can influence responding (for instance, item order), is how early items in a survery can influence responses to later items. Early items create a context within which later items are interpreted.
Well, I recently found another AR at the back of the safe. And I still have some parts left over.