TT or not TT, VAP is the question.
<i>A question that has never been answered on this blog is why do political scientists need to know quantitative methods?</i>
Same reason why anyone needs to know quantitative methods. When the number of observations in a study is higher than, say, 10 or so, relationships among variables are impossible to understand without using quantitative methods.
For example, what is the relationship between GDP and democracy? Qualitative methods are of very limited utility for answering this question.
In the first instance, as everyone knows, particularly in the social sciences, correlation does not equate causation. In other words, a statistical relationship does not in fact imply a real world relationship -- especially in the social sciences. Secondly, democracy is a qualitative concepts, which renders any such statistical relationship even less analytically useful.
Again, I ask why use quantitative methods in political science?
Quantitative scholars are well aware that correlation does not equal causation. Only a qually would think that "correlation doesn't equal causation" means that we cannot use quantitative methods.
Concepts are not "qualitative" or "quantitative." Methods are.
"Only a qually would think that "correlation doesn't equal causation" means that we cannot use quantitative methods."
It sort of does.
"Concepts are not "qualitative" or "quantitative.""
That is false. We can effectively look at voting, for instance, as a quantitative exercise, whereas democracy is a qualitative process.
It sort of does.
Uh, no it doesn't.
Let me put it this way, if correlation does not equal causation (especially in political science), then what utility do statistical methods have in this discipline?
"Let me put it this way, if correlation does not equal causation (especially in political science), then what utility do statistical methods have in this discipline?"
Because correlation is a useful subject for investigation, both in its own right and as part of causal inference. It may be inappropriate to conclude from an analysis that X causes Y, but it may still be interesting to ask whether the evidence is more consistent with the hypothesis that X causes Y or whether it is more consistent with the hypothesis that Z causes Y. That's why its called causal "inference." If there's an issue, its not with the methodology so much as with the way that the results are sometimes interpreted or discussed (ie, people claiming that their work actually allows them to establish a causal relation or absolutely rule out a competing hypothesis).
then what utility do statistical methods have in this discipline?
Identifying correlations and suggesting causal relationships.
<i>Identifying correlations and suggesting causal relationships.</i>
If statistical correlations only suggest causal relationships, what does determine causal relationships?
If statistical correlations only suggest causal relationships, what does determine causal relationships?
You can never observe causality. This is due to the fundamental problem of causal inference.
See Paul W. Holland, Statistics and Causal Inference, Journal of the American Statistical Association, Vol. 81, No. 396 (Dec., 1986), pp. 945-960.
"You can never observe causality. This is due to the fundamental problem of causal inference."
Even accepting this as true, it is still not an argument in favor of statistical analysis in political science.
Moreover, there is no theory of causation underlying these methods in this discipline. In the physical sciences there are the laws of physics. What is the analogue in political science that justifies the use of statistical methods. Again, I have yet to hear a compelling reason for the use of quantitative methods in political science.
Qually, methods don't have theories of causation. Methods are methods.
How do you italicize?
Even better use this.
See the list of allowed markup at the bottom. You can use all of them in the following format.
[tag] Stuff [/tag]
except use < instead of [ and > instead of ]
italics should be [em] Stuff [/em]
but blockquote is better.
[tag]Moreover, there is no theory of causation underlying these methods in this discipline[/tag]
And let me guess: cognition is the theory of causation underlying qualitative methods? I can't wait for you to say 'yes' to that one...
Hmmm...the original poster wrote: "A question that has never been answered on this blog is why do political scientists need to know quantitative methods?" Then a few posts down, perhaps the same poster wrote: "We can effectively look at voting, for instance, as a quantitative exercise..."
That person, whether the same or not, has effectively answered the question.
-From a "qually" who finds this thread tedious and pointless.
That person, whether the same or not, has effectively answered the question.
Wow, 8:09ed again.
I anticipate that the qually will respond that voting isn't political science. He'll probably allege that voting is engineering or something like that.
methods don't have theories of causation.
This conveys a complete lack of understanding of science, and particularly philosophy of science. In the physical sciences the entire purpose of methods is to analyze/understand the laws of physics. In the end, there is no method/science without a theory of causation.
That person, whether the same or not, has effectively answered the question.
Science is about determining why. A quantitative analysis of voting tells us how people voted, but not why.
I anticipate that the qually will respond that voting isn't political science.
I would put it closer to sociology, as I would practically all forms of behaviorism.
No one has yet offered me an explanation of why quantitative methods are used in political science.
Also, thanks very much for the info. on italics.
I have another question: how do you block quote? Thank you in advance for your help on this matter. It is greatly and deeply appreciated.
Also, thanks very much for the info. on italics.
I have another question: how do you block quote? Thank you in advance for your help on this matter. It is greatly and deeply appreciated.
You do it the same way as the italics but instead of <em> stuff </em> you type <blockquote> stuff </blockquote>
[friendly reminder] The questions about how to post things would best be new post and posted in the proper forum. You will get help faster that way and it keeps the thread on topic. [end friendly reminder]
If this thread topic is the kind of drivel that the new format is going to encourage, then I vastly prefer the old one.
This is the same type of thing that was on the old blog. The beauty of this format is that you never have to look at this thread or indeed any thread in the debate section. You can go directly to the rumors in your field. Hopefully the moderator will be strict with those areas and let the crap stay here.
I think part of the confusion is that people are not seeing the field forums on the lower part of the home page and think that the "Latest Post" section is the entire deal, when the Forums are actually the more important thing. Maybe the mod should consider switching the two sections so people are not so confused.
I would put it closer to sociology, as I would practically all forms of behaviorism.
This officially ends the debate. Let's move on to the next discussion, shall we?