Old Methodology/Theory Debates
posted by American and Comparative Jobs at 2/14/2008 07:50:00 AM
American and Comparative Jobs
Hello bloggers! As requested, I have turned off comment moderation. But, if it becomes a personal free-for-all I will ask someone else to manage the blog because I do not have time to devote to continuously deleting comments.
Thursday, February 14, 2008
Links
Previous Posts
- Old Methodology/Theory Debates
- Old Market Advice Dec. 15>
- Old Comparative Job Rumors Dec. 15>
- Old American Job Rumors Dec. 15>
- Old American Job Rumors
- Old Comparative Job Rumors--October 25th>
- Old Market Advice
- Old Methodology/Theory Debates
- Old American Job Rumors October 2007>
- Old Comparative Job Rumors
Contributors
646 Comments:
Looking at Rep. Wayne Gilchrest's website there is no mention of any anti-war positions. As a matter of fact, on the first page of the site is a picture of Gilchrest with General David Petraeus -- the architect of the "surge."
On Walter Jones's website there is no mention of any opposition to, nor criticism of, the Iraq War.
On Ric Keller's website, not only is there no vestige of opposition to the War, but there is video, on the first page, detailing Keller's travels to Iraq. This video is explicitly pro-occupation.
So much for anti-War Republicans.
In the case of Hastert (formerly Speaker), his retirement was directly linked to the loss of the House to the Democrats. Nonetheless, the loss of his seat to the Democrats was somewhat of a shock. Indeed, the loss of the Senate majority to the Democrats in 2006 was very surprising -- specifically, because so few Republican Senate seats were up for reelection.
"Still waiting for you to explain why qual. methods are the only legitimate methodology?"
That has never been asserted.
Me: "Still waiting for you to explain why qual. methods are the only legitimate methodology?"
You: "That has never been asserted."
Then please tell me the meaning of the paragraph quoted below from 3/09/2008 7:31 AM:
"Hence, any legitimate methodology would focus on human cognition to understand what prompts human social/political behavior. Any methodology that focuses on anything else to analysis human social/political behavior is less than scientific."
You seem to say one thing, then assert another, then yet another. I really do not know where you stand anymore.
Quantitative methods, for instance, can be used in conjunction with qualitative methods. (That has been asserted in the past.) I also cannot preclude the possibility that quantitative methods can be devised to meaningfully study human cognition.
"So much for anti-War Republicans."
From an article in Reason Magazine (May 8, 2007) in which Gilchrest is interviewed:
"Within two years of that vote, after taking multiple trips to Iraq, Gilchrest repudiated the decision. Talking to him the day after President Bush vetoed the Iraq supplemental funding bill—Gilchrest and Jones were the only Republicans who voted to override that—it seems incredible that he could have ever supported the war."
Moreover, bringing in Jones:
"That's what Walter and I were just working on (a document titled "Finding a Bi-Partisan Diplomatic Solution to End the War in Iraq.")," Gilchrest explains. He and Rep. Walter Jones (R-N.C.)—the North Carolina Freedom Fryer turned anti-war tubthumper—have been brainstorming ways to invigorate the Iraq debate."
Now, to bring in Keller, who came under fire by pro-war republicans for his opposition:
"How do you interpret the Republican base on this issue? There are a number of ad hoc groups that bloggers have started to punish Republicans who've cast anti-war votes, like Florida's Ric Keller..."
Sometimes we need to move beyond looking at their websites.
"Sometimes we need to move beyond looking at their websites."
We are talking about the way these MC's are casting themselves in the run up to the November election. The websites are excellent indicators of this. Going by this indicator, these candidates are not distancing themselves from the War -- as the "election above all else" hypothesis would have us believe. Indeed, two have on the first page of their websites expressly pro-War messages.
"Going by this indicator, these candidates are not distancing themselves from the War -- as the "election above all else" hypothesis would have us believe."
The point is that they do not need to completely disavow the war. They merely need to show that they are trying to find a way to end it as their constituents would prefer, while at the same time, not looking weak like their democratic counterparts (i.e., a complete and immediate withdrawl).
"Indeed, two have on the first page of their websites expressly pro-War messages."
For Jones: I do not see "no mention of any opposition to, nor criticism of" to indicate a pro-war stance. There is also apparently no mention of support either is there?
For Keller: as the article points out, he came under huge fire (and almost lost his reelection bid) due to his opposition. Thus, I think it is perfectly consistent with the office retention goal that he makes some more supportive statements.
It makes sense that there's no mention of anti-war positions on Gilchrest's website. He was heading into a very tough primary (that he lost) with two more conservative, pro-war Republicans. Last I heard he was considering endorsing the nominal Democratic challenger. It could have been that he was trying to posture himself come November by taking the "anti-war" position...But it's not likely - that's a pretty safe Republican seat he held for awhile. I think the guy's just an anomaly.
I would not trust the slant of a _Reason_ article. It is notoriously right-wing.
I find the websites more convincing of how these candidates are casting themselves. Moreover, with the unpopularity of the War why would not Jones embrace an anti-war stance if his prime goal is reelection.
Same issue with Gilchrest. The last poll released (granted it was by his opponent) had Jones in a dead heat with financial adviser Joe McLaughlin in the Republican primary. But again, like Gilchrest, Jones has had a solid lock on a pretty conservative district for a long time. I don't see how his opposition to the war is posturing for November.
I would not trust the slant of a _Reason_ article. It is notoriously right-wing.
-----
I assume that you'd trust a Mother Jones article, though, because it's so fair and balanced?
"I assume that you'd trust a Mother Jones article, though, because it's so fair and balanced?"
Who has asserted that?
"Same issue with Gilchrest. The last poll released (granted it was by his opponent) had Jones in a dead heat with financial adviser Joe McLaughlin in the Republican primary. But again, like Gilchrest, Jones has had a solid lock on a pretty conservative district for a long time. I don't see how his opposition to the war is posturing for November."
What does this mean with regard to the "election above all else" assumption?
"What does this mean with regard to the "election above all else" assumption?"
I actually didn't catch the context the assumption was used, so I'm not entirely certain. I do think that by staking out the positions they did on the war, Gilchrest and Jones made themselves more vulnerable electorally. These guys had been winning generals 70-30 the last couple go-rounds. It cost Jones a ranking member spot as well.
I think in both cases these are simply sincere positions that are tough to capture with macro-level theory. It happens.
Keller's got a tighter general election to worry about. He's probably in a different boat.
Reason has as of late run notoriously anti-war pieces, and an interview with the notoriously self-described "liberal to radical" co-creator of "The Wire," Ed Burns on the failure of the drug war, the Iraq War, public education in the US, etc.
I have heard many Republican MCs either repudiate the war altogether, or at least criticize Bush's handling of it vehemently (in a stronger sense that McCain's "we needed more troops from the beginning":
Ron Paul (d'uh)
Chuck Hagel
John Warner
Susan Collins
Norm Coleman
Tom Coburn
John Duncan
But all this is besides the point. I love how, once again, Mr. ILP brings up impeachment and the overwhelming "support" for it among the public. How does he assert this premise? Has he spoken with all U.S. voters and non-voters to figure out their cognition?
Or does he rely on opinion surveys that attempt to generalize to a broader population based on a sample? So you argue that quantitative methods are of limited use based on a premise that emanates from large-N surveys? Strange way to criticize quantitative work that may rely on... large-N surveys...
"Or does he rely on opinion surveys that attempt to generalize to a broader population based on a sample? So you argue that quantitative methods are of limited use based on a premise that emanates from large-N surveys? Strange way to criticize quantitative work that may rely on... large-N surveys..."
Voter surveys are not large-n quantitative work.
What about generalizing the opinions of 200,000,000 adults from a survey of 1,000 people?
"What about generalizing the opinions of 200,000,000 adults from a survey of 1,000 people?"
It is not quantitative analytical research.
"I think in both cases these are simply sincere positions that are tough to capture with macro-level theory."
So the reelection "above all else" hypothesis cannot verified by examining the Republican Party and the macro behavior of Republican MCs, nor can it verified on the micro level (i.e., by looking to individual MCs). How can this supposition be falsified?
12:52 here...I think I missed the debate. I really have no idea what the reelection 'above all else' claim is referring too. Just trying to provide some background on Gilchrest, Jones and Keller.
Do you think you can generalize opinions to 200,000,000 adults based on a sample of 1,000 adults? I'm not asking you whether you think this is analytical research, I'm asking whether you believe that we can generalize opinions based on surveys.
I'm still trying to understand where the premise that "Americans were for impeachment" comes from. Were you able to speak to every American? If not, how can you claim this? On what basis?
Do you think we can generalize opinions to 200,000,000 adults based on 1,000 people?
Your reluctance to answer the questions speaks volumes, by the way...
There are two notions of generalization at work here. There is descriptive generalization, and there is analytical generalization (i.e., causality). We can, for instance, descriptively generalize about the features of a species of animal -- e.g., lions have tails, tigers have four legs, etc. Then there is the question of why these features features develop. This relates to analytical generalization. To questions like these we look to the laws of physics, and evolutionary theory.
In the case of surveys, we are looking to collect description data. What do people think? (In other words, we are describing what people think.) But not analytical reasons. Why do people think this? To address this latter issue, we must delve into human cognition.
Someone asked earlier for positive proof that human behavior is driven by cognition. That is like asking to be shown gravity, or the law of conservation. While we cannot see these "laws", we know they exist by observing the behavior of matter. Similarly, while we cannot see cognition (at least not yet), we know that human behavior is determined by it.
This is why the instinct to quote Descartes was understandable: "I think therefore I am." At the heart of human behavior is thinking. While we can/should question everything else, we cannot question that.
"Someone asked earlier for positive proof that human behavior is driven by cognition."
Please see the question posted in 3/10/2008 8:01 AM. The question was why this means that qual. methods are the only legitimate methodology?
As explained in 8:07, "Quantitative methods . . . can be used in conjunction with qualitative methods."
This is why the instinct to quote Descartes was understandable: "I think therefore I am."
Have you read Descartes in French? If not, how do you know that the version you read isn't contaminated by the "cognition" of the person who translated his work? Are you sure that you understand Descartes' "cognition"?
"Je pense, donc je suis..." :)
CAUSAL CLAIM #1 on 3/09/2008 12:27 PM:
"On this question of MC reelection, we can see that in 2006 Republicans suffered an electoral slaughter **because** of their solid support for the Iraq War."
CAUSAL CLAIM #2: 3/09/2008 5:46 PM:
"The point is the Republican Party is expecting huge losses **due** to its support of the War, and has already experienced them. Nevertheless, the Party is maintaining this position. **Given this**, it appears that the Party and its MCs are not primarily motivated by reelection."
CAUSAL CLAIM #3: 3/10/2008 7:16 AM:
"The Republican Party is not backing-off its support for the Iraq War, and **as a result** Republican MCs are expecting an election blood-bath come November. This is *why* we are seeing so many Republican MCs retire, and many doing so early in their congressional careers."
Here are at least 3 causal claims made by ILP--I've put key words in asterisks.
Now, correct me if I'm wrong, but how can we know that the Republicans lost *because of the war* or that MCs are retiring *because of the expected bloodbath*? Have you spoken with any of these players as well as voters to get at their cognition, the reasons why they are retiring, the reasons why they voted for Democrats in 2006? Seems to me like you're generalizing here based on the patterns you've observed.
3/09/2008 12:27 PM: we can see that in 2006 Republicans suffered an electoral slaughter because of their solid support for the Iraq War. Nonetheless, the Republican Party's support for the War has not diminished, and it is fully expected that will continue to suffer losses for it.
Also, you claim that based on past results, we can expect Republicans to suffer losses in 2008. How so? How can you predict future results based on past results? Seems to me like you're generalizing here--which is a no-no according to your own words. How do we know that what happened in 2006 helps us predict what will happen in 2008? Could it be that it's because there are generalizable patterns that occur in all elections, within a reasonable confidence range?
@3/11/2008 1:22 PM -- sensibly argued, but I don't know why you bother.
Expect one or more of the following responses:
a) "I couldn't follow your example"
and/or
b) "but this isn't analytical political science"
and/or
c) "but I never said quant methods couldn't be used with qualitative methods"
and/or
d) "shame on you for being...so, so...logical!"
The ILP-bot has definitely passed the Turing test: my congratulations to its creator.
I haven't seen any (and I mean ANY) textbook or article mention the premise that the use of quantitative methods ought to be predicated on the presence of immutable laws. Or not even that their utility is limited in the absence of such laws. It strikes me that ILP's argument is a true paradigm shift in the Kuhnian sense.
I look forward to seeing his argument put forth in peer-reviewed journal, as it has the potential of completely shifting the discipline. But he ought to recognize that his position is NOT at all demonstrated in the literature--hence the paradigm shift.
"Also, you claim that based on past results, we can expect Republicans to suffer losses in 2008. How so? How can you predict future results based on past results? Seems to me like you're generalizing here--which is a no-no according to your own words. How do we know that what happened in 2006 helps us predict what will happen in 2008? Could it be that it's because there are generalizable patterns that occur in all elections, within a reasonable confidence range?"
I am not predicting the 2008 outcome based on what occurred in 2006, per se. I am basing it on the poll numbers. These numbers showed that the Republicans suffered electoral defeat in 2006 mostly for their support of the War. (Yes, voters' cognition was sampled via exit polls.)
The public's opinion has not changed on this issue, and neither has the Republican MCs' position, so I reasonably conclude the Party will suffer further electoral defeat on 2008. Thus, whatever pattern is observed here is derived from an understanding of human cognition, and not independent of it.
I will let others respond to the latest post. But... WOW!
"(Yes, voters' cognition was sampled via exit polls.)"
So where are we? You are now saying that quant methods **can** measure/assess 'cognition' at least to the extent that qual methods can?
FWIW, I don't think *anyone* can measure cognition (outside of neuro-science), but it's nice to know you think quants can do it.
Since when have public opinion surveys been considered quantitative analytical methods?
Since never.
I am not predicting the 2008 outcome based on what occurred in 2006, per se. I am basing it on the poll numbers. These numbers showed that the Republicans suffered electoral defeat in 2006 mostly for their support of the War. (Yes, voters' cognition was sampled via exit polls.)
The public's opinion has not changed on this issue, and neither has the Republican MCs' position, so I reasonably conclude the Party will suffer further electoral defeat on 2008. Thus, whatever pattern is observed here is derived from an understanding of human cognition, and not independent of it.
OK, let me get this straight. You say that:
"These numbers showed that the Republicans suffered electoral defeat in 2006 mostly for their support of the War",
Aren't you making a causal claim using quantitative data? Support for war causes defeat (X causes Y). And then you tell us that you got this via exit polls. So, in fact, you are generalizing based on a sample, and you are making what you earlier dubbed an "analytical generalization" rather than a "descriptive generalization". If you are able to generalize this causal claim from an exit poll to a population, you are doing quantitative analytical research, my friend.
Then you say that because X hasn't changed (support for the War and GOP position) then you should expect Y to occur again. So despite your claims to the contrary, you are making a claim about the future based on the past, i.e., you are generalizing from one election to another.
How do you reconcile this with your earlier positions about the limited utility of quantitative methods, as well as the impossibility of generalizing based on quant methods (since any study can be shown to be incorrect later, according to you)?
It's funny, because you say that you can "reasonably" conclude that the GOP will suffer losses. Earlier in the discussion, you had stated that we cannot know whether a finding with respect to one election would also be found in a later election (since studies contradict themselves frequently). What gives?
Your position keeps changing with every post...
"Then you say that because X hasn't changed (support for the War and GOP position) then you should expect Y to occur again. So despite your claims to the contrary, you are making a claim about the future based on the past, i.e., you are generalizing from one election to another."
Again, I would submit that in a scientific sense, generalization does not exist in the social sciences. In light of this, the question I would pose is does all anti-war sentiment manifest itself in voting surges against the Republican Party? Of course it does not. Evidence the Korea and Vietnam wars, where anti-war public opinion resulted in Republican presidential victories.
This is really all a question of thinking scientifically. My broad point is that too many quantoids do not. Hence, the seeming inability of posters on this blog to understand the appropriate role of quantitative methods in the social sciences -- as well as there limitations in this context.
My position has been erroneously interpreted as anti-quantitative methods. When, in truth, my position is pro-science.
And as a qualtoid, you clearly do not think very broadly. First off, no one said that all anti-war sentiment leads to GOP losses. Perhaps you could provide a citation or a reference to that effect, but I'm not aware of any literature that states this relationship. So that's a straw man.
You brush away all the criticism under the guise that you're interested in "science". D'uh! That's not a terribly convincing argument. But what you've shown is that in order to learn about "science" in the case of war and GOP success, you make a causal claim using quantitative data. That's all I'm saying.
My position has been erroneously interpreted as anti-quantitative methods. When, in truth, my position is pro-science.
That's funny, really. Because when you say that quantitative methods have limited utility in uncovering cognition, and you equate cognition with science, then what does that say about your position on quantitative methods?
I'll reiterate that you're engaging in intellectually dishonest hand-waving here. I brought up specific questions, to which you answer vaguely about "all anti-war sentiments leading to GOP losses", the Korean War, the Vietnam War, and, finally, "science."
Too bad you're unable to answer the questions asked of you. You wouldn't last a minute in a Q&A at an R1 department...
"And as a qualtoid, you clearly do not think very broadly. First off, no one said that all anti-war sentiment leads to GOP losses."
A prime shortcoming of quantoid thinking is that it tends to be too broad. Hence, my claim that generalization in a scientific sense does not exist in the social sciences. Thus, I was discussing two specific elections, and not generalizing at all. If I were truly generalizing about Republicans, elections, and war, I would have made a claim like "anti-war sentiment accrues against Republican candidates during elections."
My specific point is that you misuse/misunderstand the concept generalization. This is why I hold that you and many quantoids do not comprehend science. While in contrast, I view myself as the defender and champion of science on this blog. You are correct that this is a rather banal position, but one I feel I have to declare given the ignorance of science of too many posters on this blog.
You wrote: "you make a causal claim using quantitative data."
Qualitative scholars use quantitative data all the time. What they tend not to use are quantitative analytical methods. There is a vast difference.
"You wouldn't last a minute in a Q&A at an R1 department..."
You have no idea what is the ranking and status of my department. So I ask you not to make references to it.
Thus, I was discussing two specific elections, and not generalizing at all
Nope, you were generalizing. You misunderstand/misuse the concept of generalization. I suggest you brush up on the meaning.
And unfortunately for you, saying that you understand science when most of your postings suggest otherwise isn't terribly convincing either.
can you tell me on what basis you were able to conclude that the GOP suffered losses due to its position on the war? What is the evidence? T-tests? Multiple regression? What? What did you do with the exit poll data in order to come up with that conclusion. I'm interested in replicating your analysis.
Why so defensive regarding your institution? Sounds to me like you have a chip on your shoulder...
This is why I hold that you and many quantoids do not comprehend science.
I'm starting to think that Immutable Law Guy might actually be Gene Ray.
But I admit that I was educated stupid.
That, or he's Dark Helmet.
can you tell me on what basis you were able to conclude that the GOP suffered losses due to its position on the war?
On the basis of their human cognitions, you simpleton. Measured with this here Qualitative Human Cognitometer, generating data that is so amazingly innumerate that even to see it would cause a quantoid's head to implode.
"can you tell me on what basis you were able to conclude that the GOP suffered losses due to its position on the war? What is the evidence? T-tests? Multiple regression? What? What did you do with the exit poll data in order to come up with that conclusion. I'm interested in replicating your analysis."
A huge percentage of voters in exit polls rated the War as their number one issue. Those voters, by a two-to-one margin (perhaps more), voted Democratic. It was widely reported in the media.
"Nope, you were generalizing. You misunderstand/misuse the concept of generalization. I suggest you brush up on the meaning."
Given the large number of elections, how can discussing two specific elections be viewed as yielding a generalization? Below is the definition of "generalization."
Generalization: 1. The act or process or quality of being general 2. A general idea, statement, etc. resulting from this; inference applied generally.
Source: Webster's New World Dictionary
"...I would submit that in a *scientific sense*, generalization does not exist in the social sciences." [emph added]
and then, when generalization is *defined* it is lifted from...
"Source: Webster's New World Dictionary"
Hilarious! Websters, eh? Wow, very 'scientific'. I bet that's where Feynman used to look up all the stuff for his work on quantum computing.
Quite apart from this, are you *sure* there are no generalizations in the social sciences?
Try getting a DUI driving ban, and take a look at your car insurance rates. Why do they go up?
Alternatively, try defaulting on a few mortgage payments and then going to your bank manager for a loan. Why will he tell you "no way"?
A huge percentage of voters in exit polls rated the War as their number one issue. Those voters, by a two-to-one margin (perhaps more), voted Democratic. It was widely reported in the media.
OK. You use data from exit polls and use this to make an inference about the US population at large.
Then you write that generalization means
inference applied generally.
Congratulations, son. You've just generalized.
"Try getting a DUI driving ban, and take a look at your car insurance rates. Why do they go up?"
"Alternatively, try defaulting on a few mortgage payments and then going to your bank manager for a loan. Why will he tell you 'no way'?"
In the first instance, you are mistaking institutional rules for scientific generalization. Secondly, were there a lot of mortgage foreclosures in the former Soviet Union, and automobile insurance rate hikes? Given that you hold that these things are generalized, I assume that such phenomena occurred in the Middle Ages, where they did have housing and transportation -- as they did in ancient Rome.
"OK. You use data from exit polls and use this to make an inference about the US population at large."
"Then you write that generalization means"
"inference applied generally."
"Congratulations, son. You've just generalized."
You have not been following the discussion closely enough. What is being described here is descriptive generalization. In other words, "people have two legs, people think x, etc." This is to be differentiated from analytical (causal) generalization. "Why do people have two legs, and why do they believe x, etc.?"
" you are mistaking institutional rules for scientific generalization. "
I see: now, is this the Websters definition you are using, or something else? BTW: your insurance rates go up because you are likely to be another accident/do it again. It's a company procedure *based on a generalization* about behavior.
"Given that you hold that these things are generalized, I assume that such phenomena occurred in the Middle Ages, where they did have housing and transportation"
Is this the best you can do?
Your claim is that 21st century insurance (or loan) markets cannot be driven by generalizations of behavior because the didn't have car insurance in the Middle Ages?
Do you think there is no link between smoking and lung cancer because tobacco didn't arrive in Europe till the 17th Century?
Does that example help you see why your argument is somewhere on a continuum between "pathetic" and "a joke"?
"Your claim is that 21st century insurance (or loan) markets cannot be driven by generalizations of behavior because the didn't have car insurance in the Middle Ages?"
"Do you think there is no link between smoking and lung cancer because tobacco didn't arrive in Europe till the 17th Century?"
"Does that example help you see why your argument is somewhere on a continuum between 'pathetic' and 'a joke'?"
It is clear you are very much beyond your depth on these issues, so if I were you I would refrain from hurling insults.
The second issue has no application to this discussion, since it involves the laws of physics -- i.e., the impact of tobacco smoke on human lung tissue.
I am uncertain what your claim about insurance rates is.
You have not been following the discussion closely enough. What is being described here is descriptive generalization. In other words, "people have two legs, people think x, etc." This is to be differentiated from analytical (causal) generalization. "Why do people have two legs, and why do they believe x, etc.?"
Well, I'm just using YOUR definition of generalization. If you don't like it, maybe you should find a new one.
At any rate, you're still using quantitative data to make what you call an analytical generalization. You've said that Republicans lost because of anti-war sentiments. This is a causal statement. You measured anti-war sentiments through survey research. Then you measured Republican success by looking at all races and their outcomes. You claim that a causal relationship exists between the two.
You've generalized.
"At any rate, you're still using quantitative data to make what you call an analytical generalization. You've said that Republicans lost because of anti-war sentiments. This is a causal statement. You measured anti-war sentiments through survey research. Then you measured Republican success by looking at all races and their outcomes. You claim that a causal relationship exists between the two."
I am just aggregating data. Data collected in ONE election. There is no generalization here.
No, you're aggregating data across about 468 separate elections that occurred in 2006, and generalizing between them.
Also you are a bitter, unlikable loner who will not be mourned. It says so right here in your personnel file. "Unlikable. Liked by no one. A bitter, unlikable loner whose passing shall not be mourned. SHALL NOT BE MOURNED." Very formal. Very official. Also it says you were adopted, so that's funny.
"It is clear you are very much beyond your depth on these issues"
...says a man (and it is always a man, isn't it?) who quotes Webster's Dictionary as "scientific"
"I am uncertain what your claim about insurance rates is."
***BINGO!***
(3/11/2008 2:16 PM, option 'a')
BTW: if you don't understand something as simple as the insurance market, and how it uses generalization to set policy prices, I'm not sure you should be allowed to teach others about, well, anything.
"Unlikeable" -- perhaps -- perhaps proudly. A "loner" -- my partner would attest to the contrary.
So your position is the 468 or so elections 2006 had certain commonalities, and from those commonalities we can generalize across all these elections. Again, actual scientific generalization implies all elections, not just those in the U.S. in 2006.
"if you don't understand something as simple as the insurance market, and how it uses generalization to set policy prices, I'm not sure you should be allowed to teach others about, well, anything."
The question is not whether the insurance industry uses generalizations to set rate policies. The issue is how reliably do these generalizations predict the future. Any actuary will acknowledge that they infer from the past to determine future risk, and that the past is not necessarily a good indicator of the future.
"actual scientific generalization implies all elections"
No it doesn't: one can have qualifiers and still be perfectly 'scientific'.
Biologists don't always refer to "all species" when they make generalizations; physicists don't talk about "all matter" when they make generalizations etc.
I'll grant that they do sometimes make 'universal'-like statements, but much of the time they don't. It's still 'generalization' and it is still worthy of the term.
"The issue is how reliably do these generalizations predict the future."
So that's your problem? That there is an 'error term'? That systems are probabilistic rather than deterministic?
To give you just *one* example in physics, 'fluid mechanics' is all about *stochastic* modeling. Your claim is that this doesn't count as real generalization?
That'll be news to your campus physics dept.
"So that's your problem? That there is an 'error term'? That systems are probabilistic rather than deterministic?"
"To give you just *one* example in physics, 'fluid mechanics' is all about *stochastic* modeling. Your claim is that this doesn't count as real generalization?"
The problem with your thinking is that in physics if the fluid goes outside the range of predicted behavior, then that redounds negatively upon the theories used to make the predication. This will yield a recalibration of the theories play.
In terms of the social sciences, as in the insurance industry, human behavior regularly fails outside reasoned expectations. The prime factor behind this is not flawed theories, per se, but that humans are driven by one ultimately unpredictable variable: human cognition.
In other words, there is no way to reliably model human behavior. This is in sharp contrast to matter and energy.
"Biologists don't always refer to "all species" when they make generalizations; physicists don't talk about 'all matter' when they make generalizations etc."
What can be said about all U.S. elections, qualifiers included?
That should read: "This will yield a recalibration of the theories [at] play."
In other words, there is no way to reliably model human behavior. This is in sharp contrast to matter and energy.
Because as we all know, human behavior is something other than matter and energy.
But how do you go from exit poll data to the general claim that "Republicans suffered losses in 2006 due to their position on the War"?
Shouldn't your claim be that "Republicans didn't get the vote of these specific respondents because those respondents ranked the War as their number one concern and supported Democratic candidates. However, since cognition varies from person to person, I am unable to tell whether those respondents' views are consistent with the views of those who were not surveyed"?
You know, generalization isn't only temporal, it's also spatial, my friend. So if I say that "voters punished the GOP due to its position on the war" based on exit poll data, I am generalizing, from a sample of voters, to all voters. That's a generalization, whether you admit it or not.
So you actually generalized about the 2006 elections (going from the particular of survey respondents to the general electorate). But you did more than that. You used those results to predict the outcome of the 2008 election. So despite your claim that the past cannot predict the future (see your earlier post), you did just that. And it's okay, we don't think you're wrong on that one. We would like you to admit that you are engaging in social science. Somehow, you are reluctant to admit it.
By the way, have you read Descartes in French? Your silence says "no", which puts into question your interpretation of Descartes' cognition.
Or do you think that non-Spanish speakers can do research in Latin America by reading translations?
Shouldn't your claim be that 'Republicans didn't get the vote of these specific respondents because those respondents ranked the War as their number one concern and supported Democratic candidates. However, since cognition varies from person to person, I am unable to tell whether those respondents' views are consistent with the views of those who were not surveyed'?
You are correct that cognition varies, even in terms of why opponents of the War voted against the Republican -- or, perhaps, for the Democrats. Nonetheless, the fact remains that voters seemingly have to vote either for the Democrats or Republicans. So regardless of the precise quality of each individual's cognition, hostility to the War was translated into a vote for the Democrats.
"So you actually generalized about the 2006 elections (going from the particular of survey respondents to the general electorate)."
There are so-called sampling laws that allow for this generalizing. This, of course, reinforces the point that generalization requires "laws".
"You used those results to predict the outcome of the 2008 election."
It is not true that I used the 2006 election to predict the 2008 outcome. That would be inductive reasoning -- generalizing from the specific. What I am doing is utilizing current polling data, and party strategy, to predict that 2008 will be a replay of 2006. Thus, what I am actually saying is that voter cognition remains the same as 2006, and the Republican Party remains the same. Based on this, I am making a prediction.
"So despite your claim that the past cannot predict the future (see your earlier post), you did just that."
I never denied that we can make predictions. It would be foolish to suggest otherwise. The issue is how reliable are those predictions. Because of the immutable laws of physics we can reliably predict the behavior of energy and matter. Because of human cognition we cannot reliably predict human behavior.
If so, gambling on sporting events, for instance, would be mute. So there is no way I am betting against water at sea level boiling at 212 F, but I may bet against the Eagles minus 7 points.
" humans are driven by one ultimately unpredictable variable: human cognition."
You mean, I assume, that we cannot (yet) predict it with 100 percent accuracy.
That is, there is a *probabilistic* component to behavior: like fluid dynamics, then?
Certainly you don't really believe your statement in a deterministic sense: human behavior (political or otherwise) *is* predictable to differing extents.
Otherwise, as others have pointed out, life would be very difficult: you'd never know, for example, if your departmental secretary was going scratch your car up with her keys in the morning etc etc because human behavior is so 'unpredictable'
So, now, once again, what are you claiming?
"You mean, I assume, that we cannot (yet) predict it with 100 percent accuracy."
"That is, there is a *probabilistic* component to behavior: like fluid dynamics, then?"
No! A theory will hold that there is 0 percent possibility that fluid will behave in "x" manner. If x does occur, then we have to rework our theories of the fluid.
Now, humans are, of course, constrained within the laws of physics. Nonetheless, within those laws there is nothing that humans cannot do. You give the example of the department secretary keying my car. What if I told you that that actually happened. Would you reply that "this is impossible?" Or would you reply that something like is unexpected, but certainly possible. Others may chime in that they have heard of similar incidents.
In contrast, if I told you that water in my house, which is at sea level, did not boil at 212 degree F, you would accuse me either of lying or adding a factor to the experiment that effected the outcome. This because we have a certain understanding of the laws of physics, and a certain set of expectations as to the behavior of matter and energy as a result.
That should read: "This because we have a certain understanding of the laws of physics, and a [definite] set of expectations as to the behavior of matter and energy as a result."
"If x does occur, then we have to rework our theories of the fluid."
Two responses:
1) fine, so let's rework our theories of human behavior: that's what theorists (of any discipline) do, everyday, all day.
2) this logic is not quite right: I have a (probabilstic) theory about what the weather will be on July 4 in Las Vegas this year. I'm guessing gales and snow are pretty low probability events, but not impossible.
There: I've made a prediction, based on a theory (about weather), and it might not be correct. There's a quite a big error term and system is very hard to predict in general. Got a problem with that?
Now let's do the same thing for political behavior...
The analogy between weather and human behavior has been used on this blog before, and it is a false one. Weather is driven by the laws of physics. This means that as we gain greater and greater comprehension of these laws our ability to predict the weather will improve -- provided our measuring instruments correspondingly improve.
Human behavior is driven by cognition. Human cognition is autonomous, creative, and, ultimately, unpredictable. So we can pose the same exact situation to 50 different individuals, and they could respond 50 different ways.
Interestingly, the inventors of game theory initially developed it to model economic consumption. After extensive efforts, they found that it was literally impossible to model how people would use their money.
"Human cognition...and, ultimately, unpredictable."
It's a like a broken record: try again, on *what* do you base this statement?
You mean "is not currently predictable with 100% accuracy". No one would disagree.
Now, why does this imply we should give up trying to predict it?
I'm quite sure cavemen thought the apperance of Halley's comet was 'unpredictable' but we know that to be untrue (now). We started coming up with theories and measuring things and worked it out.
Why are you so certain we cannot (and never will be able to) do the same for human behavior?
Please try and answer rather than stating "because human behavior is unpredictable" or the like.
If that is your irrational faith (and in the face of overwhelming evidence that's all it can be) then there really is no point debating, is there?
"If that is your irrational faith (and in the face of overwhelming evidence that's all it can be)"
What evidence are you speaking of?
"If that is your irrational faith (and in the face of overwhelming evidence that's all it can be)"
"What evidence are you speaking of?"
Err...the fact that human behavior *is* predictable: for example, that your secretary isn't going to greet you naked tomorrow morning, address you only in Russian, and then hit you round the face with a wet fish.
Yes, yes, it *could* physically occur, but you know it is pretty unlikely. You wouldn't go to work if you predicted otherwise.
So, that example, and the million other human/social interactions that you relied on going 'predictably' today (I also doubt, for example, that the mail man painted your house green while you were work).
Given that your day-to-day life is (to some extent) predictable, I'd start there.
Again, you posit a notion of predictability that is not scientific. In the first instance, you acknowledge in your post that the behaviors you suggest are possible.
Again, this is radically different in the physical science, where prediction tends to be much more certain. In other words, if you define the parameters, I can in most case predict with certainty the behavior of matter and energy. Additionally, the predictability you allude to is derived from norms and institutions. Hence, you cannot even weakly predict across societies and institutions.
Weather is driven by the laws of physics.
...
Human behavior is driven by cognition. Human cognition is autonomous, creative, and, ultimately, unpredictable.
These statements make no sense unless you believe that human cognition is supernatural.
Last I checked. my brain is made of plain dumb matter, and all my cognition boils down to matter rearranging itself using energy. By your own statements, then, my behavior can be modeled with complete certainty. That, or cognition is some sort of mystical process that lies beyond physics.
I'm assuming here that you're ignorant of ideas like Penrose's conjecture that cognition depends on collapsing quantum states and so might have a tiny core of irreducible uncertainty. Even then, the probability density of cognition could be perfectly understood and described.
Incidentally, have you tried peddling your idea that cognition is unpredictable to the cognitive-psych people? They're wasting an awful lot of time and energy explaining the unexplainable and predicting the unpredictable. I guess they've just been lucky so far.
The cognition poster here is a sad character. From a scientific standpoint, in fact, we can confidently say that without quantitative methods, political science would be little more than a set of stories. Interesting, but more akin to journalism than to social science.
Social science is fundamentally about uncovering and explaining regularities in the social world. Qualitative methods, by their very nature, are wholly unsuited for this task. They do not demonstrate causality; in fact, due to what philosophers of causation call the "fundamental problem of causal inference," qualitative methods are useless for investigating causality.
At the very best, qualitative methods can help us to put an individual observation in its context. Oftentimes this is interesting. Sometimes this helps us to hypothesize about new regularities in the social world. But without turning to quantitative methods to test whether hypothesized relationships and regularities exist in the social world, qualitative methods are not social science.
"Social science is fundamentally about uncovering and explaining regularities in the social world."
Science is about understanding causation -- i.e., determining why. Why is it about uncovering regularities in the social world?
Moreover, you are assuming there are regularities, and that they are more worthy of study then events that happen with less regularity, or are unique.
Even the very notion of regularity in terms of human behavior is rather concocted. Let us take the cases of the U.S. Civil War, World War II, and/or the Vietnam War. The idea that these are "wars" is a social construct, not a scientific nomenclature. Thus, a lot of what passes for regularity is a social construct invoked to simplify language and/or political/historical discussion.
"From a scientific standpoint, in fact, we can confidently say that without quantitative methods, political science would be little more than a set of stories. Interesting, but more akin to journalism than to social science."
This claim has no basis in reality. It is a baseless assertion about qualitative methods.
"At the very best, qualitative methods can help us to put an individual observation in its context."
This has no meaning.
"But without turning to quantitative methods to test whether hypothesized relationships and regularities exist in the social world."
Quantitative methods can uncover patterns of behavior, but cannot discern causation.
Science is about understanding causation -- i.e., determining why. Why is it about uncovering regularities in the social world?
Because that is how knowledge accumulates. We use quantitative methods to investigate if the regularities we hypothesize to exist actually do.
Moreover, you are assuming there are regularities, and that they are more worthy of study then events that happen with less regularity, or are unique.
Regularities are the hallmarks of the social world. With few exceptions, people stop at stoplights. With few exceptions, humans procreate. With few exceptions, wars kill people. To hypothesize that these are not regularities of the social world would be naive.
Quantitative methods can uncover patterns of behavior, but cannot discern causation.
In fact, there is little disagreement among philosophers of social science that because of the "fundamental problem of causal inference," causation can never be conclusively demonstrated. However, the most appropriate tools for approximating causation are experiments and quasi-experiments. Both are quantitative methods. Qualitative methods are simply incapable of demonstrating causation.
The stories that qualitative scholars tell us may be interesting, and they may prompt us to look for new quantitative data, but they do not demonstrate causation. This is widely accepted.
"Regularities are the hallmarks of the social world."
I am uncertain what this means, nor why it carries any importance with you.
"With few exceptions, people stop at stoplights. With few exceptions, humans procreate. With few exceptions, wars kill people."
Again, I do not know what this means, nor why you are invoking it. It should also be noted that lots of things kill people, including disease, spousal abuse, bank robbers, etc. Why do you raise war in particular?
"The stories that qualitative scholars cannot . . . demonstrate causation."
The prime methods that demonstrate causation in the social science are qualitative. Qualitative methods attempt to gauge the central cause of human behavior: cognition (i.e., thinking).
"I am uncertain what this means, nor why it carries any importance with you."
Really? I follow this poster quite clearly?
that should read
...clearly.
What does hallmark mean in this context?
"With few exceptions, people stop at stoplights. With few exceptions, humans procreate. With few exceptions, wars kill people."
These events happen frequently, but not seemingly with regularity. Regularity implies at regular time intervals.
"This is widely accepted."
I used to be widely accepted that the world is flat.
That should read
It used to be . . .
The quallies that post on this blog have yet to cite a single study that explains why qualitative methods can uncover causality. Moreover, they are unable to even cite a single philosopher of social science or causality who agrees with them.
Their silence speaks volumes.
Meanwhile, for examples of the logic of causal inference and the utility of quantitative methods, we can look to Neyman (1923), Rubin (1974), Holland (1986), Sekhon (2004), Rosenbaum (2002), and Bowers and Hansen (2005) as just a brief sampling. For inspiration, these works goes back to J.S. Mill.
Qualitative methods alone do not demonstrate causality. Rather, qualitative methods are at best a supplement to quantitative methods.
I used to be widely accepted that the world is flat.
What's your point? That long ago people were wrong about astronomy, so that means that philosophers of science are wrong about causality today?
The point is invoking what "is widely believed", or citations (alla 8:10 p.m.), is unconvincing, nor demonstrates anything.
The point is invoking what "is widely believed", or citations (alla 8:10 p.m.), is unconvincing, nor demonstrates anything.
By the same token, what is widely believed is not wrong because it is widely believed. What is incumbent upon the quallies is to explain why the Neyman-Rubin-Holland causal framework, the most influential understanding of causality in the philosophy of social science for the past century, is incorrect. The quallies cannot, so they are left with lame critiques with no substance. Like yours. "Well, people used to believe the earth was flat!" Please.
By contrast, quantitative methods are linked directly with a large and well-developed philosophy of causality and inference, one inspired by J.S. Mill, and elaborated since by philosophers, statisticians, sociologists, epidemiologists, economists, and political scientists.
Qualitative methods offer nothing comparable.
"I am uncertain what this means, nor why it carries any importance with you."
"Really? I follow this poster quite clearly."
Yeah, so do I.
I've followed this thread for a while, and this qual poster's strategy seems to run as follows: "if my argument is shown to be complete nonsense and I have once again embarrassed myself, I will claim to 'not understand' the example/point being presented."
It's option 'a' of 3/11/2008 2:16 PM.
Of course, this might *not* be a strategy, and the poster might *really* not understand what is being explained so patiently to him, but that is so damning of his intelligence (and the discipline I share with him) that I try to put a more optimistic spin on things.
"Qualitative methods alone do not demonstrate causality. Rather, qualitative methods are at best a supplement to quantitative methods."
I largely agree with the first sentence here. Qualitative approaches generally do not make for very convincing causal arguments. Neither do most quantitative applications with observational data, but that is another issue.
But the second sentence is highly problematic without some kind of qualifier. There is no reason to believe that demonstrations of causal relationships are the "be all and end all" of political science. Work that is rigorous but mostly descriptive is instrumental to many "hard" sciences and particularly important in investigations of social phenomena. Qualitative methods have a huge role to play there, as do quantitative ones in many applications. We've got a big tent in political science, and that's a good thing.
But the second sentence is highly problematic without some kind of qualifier.
I agree, and apologize for not providing that qualifier in the first place. I am all in favor of thick description, and think that there are important things to be learned from it.
I should have written, "Rather, for demonstrating causality, qualitative methods are at best a supplement to quantitative methods."
3/14/2008 5:43 AM
While I disagree with most of the quantoid posters on this blog, I will definitely credit them with putting forward substantive arguments. Your post, however, falls short of any substance. Your statement that "what is incumbent upon the quallies is to explain why the Neyman-Rubin-Holland causal framework, the most influential understanding of causality in the philosophy of social science for the past century, is incorrect" is just pollution.
Similarly, if I were to point out that the most influential and important works in the American subfield were qualitative (i.e., Dahl's _Who Govern's_, Lowi's _The End of Liberalism_, and Heclo's "Issue Networks and the Executive Establishment"), that would analytically accomplish nothing. So let us please refrain from devolving into dueling citations.
"Qualitative methods alone do not demonstrate causality. Rather, qualitative methods are at best a supplement to quantitative methods."
By definition quantitative methods in the social sciences cannot establish causation. All such methods can do is document patterns of human behavior.
Qualitative methods, in contrast, seek to directly measure human cognition. In other words, qualitative scholars seek to directly examine the thinking of actors when they undertake political actions. It is this thinking that causes all political actions.
"Qualitative methods alone do not demonstrate causality."
Why?
1:58 p.m.
Please do not ask questions that quant-heads cannot answer. It is embarrassing to quants, and we have too strong a pedigree to merit that.
While I disagree with most of the quantoid posters on this blog, I will definitely credit them with putting forward substantive arguments. Your post, however, falls short of any substance. Your statement that "what is incumbent upon the quallies is to explain why the Neyman-Rubin-Holland causal framework, the most influential understanding of causality in the philosophy of social science for the past century, is incorrect" is just pollution.
A laughable attempt to dissemble by a frustrated qually. I discuss the dominant framework for understanding causal relationships, one inspired by Mill and with a pedigree of nearly a century, and you say it doesn't count.
Do you even understand it?
By definition quantitative methods in the social sciences cannot establish causation. All such methods can do is document patterns of human behavior.
Without documenting patterns and variation, causation can never be proven. This has been shown on this blog, and the relevant citations provided.
Qualitative methods do not do this. Hence, they do not measure causation. Massive fail.
"Without documenting patterns and variation, causation can never be proven."
Why?
This bloggers makes bald assertions without supporting argumentation, much less evidence.
This bloggers makes bald assertions without supporting argumentation, much less evidence.
I cited half a dozen studies, and you called them pollution. Read them. They are the evidence.
"I discuss the dominant framework for understanding causal relationships."
Where?
Just saying there is a dominant framework does not mean there is one, nor does it mean this framework is appropriate. Again, the dominant framework in geography used to be predicated on the notion that the world is flat.
This argumentation through grand statements is starting to look amateurish. Moreover, I am coming to the conclusion that there is no substance underlying these sweeping claims.
"I cited half a dozen studies, and you called them pollution."
If your goal is simply to invoke citations, you are wasting your time.
Where?
3/13/2008 8:10 PM
I am starting to believe that you are illiterate.
Just saying there is a dominant framework does not mean there is one, nor does it mean this framework is appropriate. Again, the dominant framework in geography used to be predicated on the notion that the world is flat.
This argumentation through grand statements is starting to look amateurish. Moreover, I am coming to the conclusion that there is no substance underlying these sweeping claims.
Your bald refusal to engage in any point of substantive debate reveals you to be insecure. Your claim is that the dominant framework for thinking about causality for the past century is wrong, yet
1. you do not know what it is
2. you do not understand it, and
3. your evidence is that it's the dominant framework!
What is clear is that qualitative methods training has failed you in your quest to become a social scientist. This is the sad life of a qually.
RE: 2:24 p.m.
Is that how you conduct your pedagogy? You walk into class, invoke literature, and walk out. It sounds like a fascinating course! =-)
if I were to point out that the most influential and important works in the American subfield were qualitative
...I would have to start by demonstrating the influence and importance using some agreeable metric. Instead, I'll list some things that I assert are the most important works in American with no supporting evidence whatsoever.
Is that how you conduct your pedagogy? You walk into class, invoke literature, and walk out. It sounds like a fascinating course! =-)
I am not teaching a class here, I am simply asking the poster to read the literature. Not only does s/he refuse to do so, s/he apparently cannot even read the thread well enough to find them!
Qualitative methods, in contrast, seek to directly measure human cognition
Oh, so when you say "qualitative methods," you mean "fMRI and PET."
"Your bald refusal to engage in any point of substantive debate reveals you to be insecure."
How can I engage in a substantive debate with you if you refuse to put forward any substantive positions?
This is like a bad Abbott and Costello routine! (lol)
"This argumentation through grand statements is starting to look amateurish."
I don't think that the grand statements are your problem. Your problem is that the poster is correct, and this scares you.
"I am simply asking the poster to read the literature"
If your point is to exhort me to read some particular literature, you are wasting our time. Other than your assertion that this is important literature, you have not given me any reason to do so.
"I would have to start by demonstrating the influence and importance [of this literature] using some agreeable metric"
I wish the quantoid poster would abide by this approach.
How can I engage in a substantive debate with you if you refuse to put forward any substantive positions?
I did so. I cited the Neyman-Rubin-Holland causal framework. You ignore it and you say it doesn't count because everyone believes it.
If your point is to exhort me to read some particular literature, you are wasting our time. Other than your assertion that this is important literature, you have not given me any reason to do so.
Got it. You don't want to read the literature that disagrees with you. Way to go!
"I did so. I cited the Neyman-Rubin-Holland causal framework. You ignore it and you say it doesn't count because everyone believes it."
This is not a substantive position. It is solely hand-waving.
This is not a substantive position. It is solely hand-waving.
Here's to you, Mr. "Ignore All Literature That Disagrees With Me."
"You don't want to read the literature that disagrees with you."
I never wrote that. What I noted is I am not being given any reason to read this literature, except that some anonymous blogger has invoked it.
You guys are obviously not serious about engaging in a discussion about methodology. Why did you choose to post here?
What the qualitative methods poster has shown us here is the danger of qualitative methods in our discipline. While I cannot demonstrate the existence of a causal relationship based on him/her alone, these posts reveal a stunning lack of intellectual curiosity, and a dangerous inability to grasp the fundamentals of the philosophy of science from J.S. Mill until today.
"I cannot demonstrate the existence of a causal relationship based on him/her alone."
What does this mean?
3/14/2008 2:47 PM
What could the quantitative methods poster tell you that would make you read up on the philosophy of science?
"What the qualitative methods poster has shown us here is the danger of qualitative methods in our discipline."
The only danger I see in qualitative methods is that they show the folly inherent in many quantitative studies.
What does this mean?
Are you illiterate? It means that I cannot demonstrate a link between qualitative methods and inability to understand basic philosophy of science based on your posts alone. I need far more data to do this...I need quantitative data from a large number of qualitative posters to accomplish this task.
3/14/2008 2:48 PM
I agree. I can't believe that this person is a student in an American university! If one of my graduate students refused to read the literature because he didn't want to, I would have him kicked out of the program.
"What could the quantitative methods poster tell you that would make you read up on the philosophy of science?"
I know plenty of philosophy of science, something this poster has yet to demonstrate any knowledge of. If he/she could show some knowledge of philosophy of science, and the merits of his/her position, then I would be more compelled to read the specific literature he/she is pointing to.
"It means that I cannot demonstrate a link between qualitative methods and inability to understand basic philosophy of science based on your posts alone. I need far more data to do this...I need quantitative data from a large number of qualitative posters to accomplish this task."
I mean no disrespect, but this is incoherent.
"What does this mean?"
I would guess it means that the poster cannot use the single case of *you* as evidence that a qualitative orientation generally causes (or is caused by) a lack of understanding of the fundamental principles of inference.
Just a guess though.
"I mean no disrespect, but this is incoherent."
No it isn't: I understood his point.
Anyway, you said
"I know plenty of philosophy of science"
yet you are ignorant of Neyman-Rubin-Holland?
These things aren't compatible! You've never heard of
Jerzy Neyman? Seriously?
Check him out:
http://tinyurl.com/2n57bv
I mean no disrespect, but this is incoherent.
My sentence really isn't hard. If you cannot understand it, then you are probably not a political scientist.
I know plenty of philosophy of science, something this poster has yet to demonstrate any knowledge of. If he/she could show some knowledge of philosophy of science, and the merits of his/her position, then I would be more compelled to read the specific literature he/she is pointing to.
Apparently you don't. If you've never heard of Neyman, or Rubin, or their work on causation, then you really don't know the philosophy of science. This alone should compel you to start reading.
"What the qualitative methods poster has shown us here is the danger of qualitative methods in our discipline."
Are qualitative scholars the Islamic extremists of political science? (ha . . . ha . . . ha . . .)
"If you've never heard of Neyman, or Rubin, or their work on causation, then you really don't know the philosophy of science."
If you say so. Can we move on to something else?
If you say so. Can we move on to something else?
Sure, but that something else had better not be an indictment of methods that you don't understand using authors that you've never heard of.
"Sure, but that something else had better not be an indictment of methods that you don't understand using authors that you've never heard of."
This statement is seemingly made by someone that has demonstrated no knowledge of methods or philosophy of science. Moreover, it is patently authoritarian. I guess the desperate quantoid is seeking to hold on to his hegemony through repressive means.
Hey, Mr. ILP:
3/09/2008 10:51 AM
“I think therefore I am” – Descartes.
Free will guided by reason (i.e., thinking) has been a pillar of Western thought since at least the Enlightenment, if not since the Ancient Greeks. John Locke, along with Deists (including Adam Smith), have held that Natural Laws shape human behavior
3/09/2008 12:27 PM:
If you cannot accept the Western philosophical tradition as sufficient citation, than I imagine there is no citation I can invoke that you will find acceptable.
3/09/2008 1:47 PM:
It is not a question of evidence. In my mind the idea that humans are driven by their own thinking is self-evident. Now, if you want to engage in a full fledged discussion of this, just read the Western canon.
So earlier in the debate, someone asked you for evidence regarding cognition. Your response was to cite Descartes, Marx, and the "Western canon". Then someone followed up with "no, please give me evidence, no just citations," and you response was "if you don't think that the Western canon is evidence, I don't know that I can convince you" (I'm paraphrasing here).
Are you now in agreement that citations do not count as evidence? If so, then you have completely discredited your own position, since the 'evidence' in favor of it came in the form of... citations!
Very clever, but unfortunately for you, previous postings do not disappear... Of course, I'm citing you here, so if citations aren't relevant, then we're engaging in a circular debate...
Fair enough. But this was subsequently posted:
"Someone asked earlier for positive proof that human behavior is driven by cognition. That is like asking to be shown gravity, or the law of conservation. While we cannot see these 'laws', we know they exist by observing the behavior of matter. Similarly, while we cannot see cognition (at least not yet), we know that human behavior is determined by it."
"This is why the instinct to quote Descartes was understandable: 'I think therefore I am.' At the heart of human behavior is thinking. While we can/should question everything else, we cannot question that."
I'm citing you here, so if citations aren't relevant, then we're engaging in a circular debate...
You're not citing me. I didn't write that. FAIL.
This statement is seemingly made by someone that has demonstrated no knowledge of methods or philosophy of science. Moreover, it is patently authoritarian. I guess the desperate quantoid is seeking to hold on to his hegemony through repressive means.
Really? No knowledge of methods or the philosophy of science? This coming from someone who rejects the predominant conception of causation from J.S. Mill to today? I'm sorry that your feelings are hurt, but that doesn't make me authoritarian or repressive. It means that just maybe you're going to learn something today.
Incidentally, we see here the sad tactic of the qually. When his arguments have been refuted, he claims that the accumulation of knowledge corresponds to domination, hegemony, or authoritarianism. In doing so, he is just hiding behind his own weaknesses. If you had EVEN ONE philosopher of social science or causality that agreed with you, you could trot him out and we could have a debate. You don't. So there's no debate to have.
"When [the qualitative scholar's] arguments have been refuted"
What is insane is that you have not put forward one argument, nor one substantive assertion! How you could possibly claim you have refuted anything?!
I guess your tactic is to follow Goebbels's lead. If you repeat yourself enough, people will believe you.
"This coming from someone who rejects the predominant conception of causation from J.S. Mill to today?"
Please describe this predominant conception of causation from J.S. Mill to today. Here is your opportunity to change minds.
I guess your tactic is to follow Goebbels's lead. If you repeat yourself enough, people will believe you.
Internet people call this "Godwin's Law." As the thread increases in length, the probability that some frustrated poster will equate his opponent to a Nazi approaches 1.
"If you had EVEN ONE philosopher of social science or causality that agreed with you, you could trot him out and we could have a debate."
The entire Western canon points to human thought as the cause of human behavior -- from Plato to Foucault. I know you are unprepared to discuss any of these thinkers. None make such absurd claims like "regularities are the hallmarks of the social world."
6:21,
So are you consciously using Goebbels's tactic or not?
I'm not the original quantitative methods poster, but let me give you a brief introduction to what he's saying. (And by the way, it's correct.)
Since Mill, the problem of causal inference has been considered one of missing data. Neyman and Rubin have formalized it. In the very simplest case, where the causal variable of interest has only two values (it exists or it does not), we define the causal effect E of an independent variable T (for treatment) on outcome Y for individual i as
E = Y_i|T=1 - Y_i|T=0.
We unfortunately can never observe this quantity, for the same individual (i) can never be both treated and not treated. So given that, the best strategy possible is to search for as many relevant comparisons as possible, employing some parametric or non-parametric technique to estimate the average association between units that have been treated and units that have not been treated.
The problem is of course one of proper comparability; thus, experiments and quasi-experiments where the conditional probability of T=[0,1] is independent of all other characteristics X of the individuals.
Qualitative methods will never accomplish this task of uncovering causality.
So are you consciously using Goebbels's tactic or not?
I am not using Goebbel's tactic at all. By the way, have you stopped hitting your wife yet?
"By the way, have you stopped hitting your wife yet?"
I am not sure what that has to do with anything. In any event, I am not married.
The entire Western canon points to human thought as the cause of human behavior
Even if that were true--and it is not, but let's concede it for the sake of argument--it still is impossible for qualitative methods to uncover causality. This is a consequence of what is known as the "fundamental problem of causal inference."
I note also that you are inherently biased towards Western thought. As if the Western canon itself is the end-all and be-all of human understanding of causality.
I note, finally, that you're wrong. The entire Western canon does not claim that human thought is the cause of human behavior.
"The entire Western canon does not claim that human thought is the cause of human behavior."
Which political philosopher(s) are you pointing to?
RE: 6:29 p.m.
Experiments do not divide quantitative and qualitative methods. Experiments can be reported via qualitative methods. Perhaps the most famous social science experiment (conducted by Stanley Milgram) was reported qualitatively. In other words, there were no quantitative analytical methods used to determine causation, nor to analyze the experiment results.
Which political philosopher(s) are you pointing to?
Laplace, to name one. The belief that "the Western canon," whatever that means, has decided for once and for all that cognition is the cause of all social outcomes, is simply false.
Perhaps the most famous social science experiment (conducted by Stanley Milgram) was reported qualitatively.
False. The conclusions were stated quantitatively (26/40 "obeyed the experimental commands fully"). Moreover, the outcomes were never stated in causal terms, because the experiment had no manipulated treatment. A search of the term "cause" yields no results that may be interpreted as Milgram attributing causality to anything of his results from the experiment. It uncovered an interesting phenomenon, and the last several paragraphs are devoted to explaining what must be done in future research, but there is no claim about causality.
Again, we find that the quallies fail at basic research. They can't even do that right.
This demonstrates yet again that the level of ignorance among devotees of qualitative methods as the exclusive tool of political science is damaging not only to their own research, but to knowledge cumulation in political science writ large. The conclusion is inescapable. Qualitative methods do not demonstrate causality. They are interesting, and I welcome their use, but as a supplements to quantitative methods in the search of causality.
I'm not the original quantitative methods poster, but let me give you a brief introduction to what he's saying. (And by the way, it's correct.)
Thank you for posting a very basic introduction to the issues at hand. I wonder if the qually will be able to respond to it. So far, s/he's ignored it. Typical.
RE: 10:24 p.m.
"Laplace, to name one."
Are you referring to Pierre-Simon Laplace? He was not a political philosopher, nor even a social scientist. He was a mathematician and astronomer.
"The belief that 'the Western canon,' whatever that means, has decided for once and for all that cognition is the cause of all social outcomes, is simply false."
If you do not know the Western canon, why are you making any claims about it?
RE: 10:41 p.m.
"False. The conclusions were stated quantitatively (26/40 'obeyed the experimental commands fully'). "
There is a vast difference between quantitative data and quantitative methods. Qualitative scholars use quantitative data all the time.
"Moreover, the outcomes were never stated in causal terms, because the experiment had no manipulated treatment."
It is true Miligram's work in this instance was a quasi-experiment. Nonetheless, there was causality at play. The independent variable was "authority", and the dependent variable the "people's capacity for violence". Thus, the experiment indicated that under the "color of authority" a significant number of "normal" people are able to carry out violence and inflict extreme pain.
The claim that only research conducted via experiment is science would leave 99 percent of all political science work outside of "science" -- including virtually all research using quantitative methods. Experiments, or quasi-experiments, are used by very, very few political scientists. Moreover, most of the subject matter that is currently under the purview of the discipline would now fall outside of it, since most of this subject matter cannot be manipulated via experimental method -- e.g., the study of Congress.
RE: 6:23 a.m.
That should read:
The independent variable was "authority", and the dependent variable "people's capacity for violence".
Are you referring to Pierre-Simon Laplace? He was not a political philosopher, nor even a social scientist. He was a mathematician and astronomer.
He is a dominant figure in the study of probability and statistics, and he wrote influential works on the nature of determinism in both the physical and natural world. He influenced literally hundreds of subsequent works on what causality means, both among statisticians and among philosophers of science.
You don't even know the Western canon.
It is true Miligram's work in this instance was a quasi-experiment. Nonetheless, there was causality at play. The independent variable was "authority", and the dependent variable the "people's capacity for violence". Thus, the experiment indicated that under the "color of authority" a significant number of "normal" people are able to carry out violence and inflict extreme pain.
No. Milgram did not vary the treatment of authority, so there was no way to discern causality--and indeed, no attempt to show it. Milgram's results, both to Milgram and his audience, were nothing more than illustrative and suggestive.
To say that there "was causality" is trivially true.
In my view what is missing from the quant. side of this discussion is a theory of causation. In other words, the experiment method is exactly that -- a method. Implicit within it is not a theory of causation. So the actual question is what is this method detecting/measuring? Is it the "laws of social/political behavior", human cognition, or something else?
Are you referring to Pierre-Simon Laplace? He was not a political philosopher, nor even a social scientist. He was a mathematician and astronomer.
From wikipedia:
"Pierre-Simon, marquis de Laplace (March 23, 1749 - March 5, 1827) was a French mathematician and astronomer"
So, I guess ILP did a quick Google search, clicked on the first link (Wikipedia), and reported on the first sentence.
Great work, Sherlock! That's how my undergrads to their research.
Of course, you could have admitted that you didn't know Laplace, but hey, that would be asking too much. Since you've already cited Webster's on many occasions as the reference for statistical terms, I guess we shouldn't be surprised that you resort to Wikipedia to make a counter-argument.
6:59 a.m.
By your own description of Laplace's work he is not someone that philosophized on human political behavior. You wrote: "He is a dominant figure in the study of probability and statistics, and he wrote influential works on the nature of determinism in both the physical and natural world."
"follow Goebbels's lead."
I see the debate has moved on, but this strikes me as an outrageous statement.
Likening someone who holds a particular (and I should add _mainstream_ view) of social science research to a Nazi is both childish and vile.
Grow. Up.
I'm actually quite shocked that the quant poster still has the decency to debate with you. Good for him: I don't know much about Rubin and/or experiments but I know for sure who is the bigger man.
RE: 7:22 a.m.
What is amazing is that you obviously know nothing of political philosophy (your invocation of Laplace showed that), and to cover up your ignorance you start insulting people. People, mind you, that have not insulted you, and are trying to engage you in a thoughtful and respectful manner. Your lack of maturity is disappointing.
7:40,
You are overreacting to the Goebbels reference. It is common to refer to him as someone that pioneered the "big lie" method.
"cover up your ignorance you start insulting people."
"You are overreacting to the Goebbels reference."
I do so hope these were not written by the same poster...
BTW: take a look at who used "the big lie" first, at to what he was referring.
It's a stupid, stupid thing to write, and it has absolutely no place on an academic thread about methodology.
By your own description of Laplace's work he is not someone that philosophized on human political behavior. You wrote: "He is a dominant figure in the study of probability and statistics, and he wrote influential works on the nature of determinism in both the physical and natural world."
So what? If you can't understand the links between philosophical determinism and human behavior, then you really don't know anything about the philosophy of science.
Popper spent years grappling with Laplace's views. Does he not count because Laplace worked on statistics?
"If you can't understand the links between philosophical determinism and human behavior"
What philosophical deterinism are you referring to?
In my view what is missing from the quant. side of this discussion is a theory of causation. In other words, the experiment method is exactly that -- a method. Implicit within it is not a theory of causation. So the actual question is what is this method detecting/measuring? Is it the "laws of social/political behavior", human cognition, or something else?
Both qualitative methods and quantitative methods are methods. Qualitative methods are just that: methods. Qualitative methods do not imply a theory of causality.
Quantitative methods do not imply a theory of causality. They, like, qualitative methods, are methods. However, quantitative analysis strives to grasp with the inferential problems made explicit in the Neyman-Rubin model. This is why they are suitable for uncovering causality, while qualitative methods--which can make no such claim--are not.
What is amazing is that you obviously know nothing of political philosophy (your invocation of Laplace showed that), and to cover up your ignorance you start insulting people. People, mind you, that have not insulted you, and are trying to engage you in a thoughtful and respectful manner. Your lack of maturity is disappointing.
All aboard the failboat!
1. I didn't write that
2. Your idea of thoughtful and respectful is to call someone a Nazi
"Qualitative methods do not imply a theory of causality."
You are wrong. Embedded within qualitative methods is the idea that human cognition (i.e., thinking) causes human political behavior. Qualitative methods seek to measure this thinking, and understand/analyze how it leads to political behavior.
"Quantitative analysis strives to grasp with the inferential problems made explicit in the Neyman-Rubin model."
I understand. But what are you trying to infer -- the laws of social/political behavior, cognition, or perhaps something else? Again, there has to be theory of causation underlying every method. Otherwise, the method is useless.
You are wrong. Embedded within qualitative methods is the idea that human cognition (i.e., thinking) causes human political behavior. Qualitative methods seek to measure this thinking, and understand/analyze how it leads to political behavior.
That is no more embedded in qualitative methods than it is in quantitative methods. I.e., you can think that, but the methods don't entail it.
I could just as easily write that embedded within quantitative methods is the idea that human cognition (i.e., thinking) causes human political behavior. Quantitative methods seek to measure this thinking, and understand/analyze how it leads to political behavior.
The difference would be, quantitative methods actually allow us to approximate the task of uncovering causality, while qualitative methods cannot.
"That is no more embedded in qualitative methods than it is in quantitative methods. I.e., you can think that, but the methods don't entail it."
I do not know why you make this claim. It is common for you to make claims without any substantiation.
"I could just as easily write that embedded within quantitative methods is the idea that human cognition (i.e., thinking) causes human political behavior. Quantitative methods seek to measure this thinking, and understand/analyze how it leads to political behavior."
If this was your position, we could engage in a discussion about the effectiveness of quantitative methods to gauge human cognition, versus the ability of qualitative methods. In other words, we can have a real discussion about methods.
"The difference would be, quantitative methods actually allow us to approximate the task of uncovering causality, while qualitative methods cannot."
In terms of "uncovering causality", I have consistently argued, as have others, that uncovering causality necessitates examining human cognition -- since this is the prime cause of political behavior. With regard to determining causality, what do you seek to examine (i.e., uncover) -- human cognition, the laws of social/political behavior, or something else? Once you answer this we can discussion the merits and demerits of quantitative and qualitative methods.
I do not know why you make this claim. It is common for you to make claims without any substantiation.
For the same reason that you made the claim that experiments do not have a theory of causality. Are you really this dense?
In terms of "uncovering causality", I have consistently argued, as have others, that uncovering causality necessitates examining human cognition -- since this is the prime cause of political behavior. With regard to determining causality, what do you seek to examine (i.e., uncover) -- human cognition, the laws of social/political behavior, or something else? Once you answer this we can discussion the merits and demerits of quantitative and qualitative methods.
Methods do not entail causal claims. We can discuss them without agree about what we think causes things.
Moreover, thinking that cognition is the cause of behavior is not the same as uncovering causality. Uncovering causality requires an understanding of the fundamental problem of causal inference. Quantitative methods approach this task. Qualitative methods do not. Moreover, you refuse to even acknowledge the problem. You do this by refusing to read the literature, and by refusing to acknowledge it when it is spelled out to you.
I interpret this as your surrender.
"For the same reason that you made the claim that experiments do not have a theory of causality. Are you really this dense?"
I see your point. The difference is that I, and others, have explained thoroughly and logically the theory of causation embedded in qualitative methods in the social sciences.
If we were discussing the physical sciences, I could immediately invoke the theory of causation underlying the experiment method: "the laws of physics cause the behavior of matter and energy." In the social sciences, where there are no such "laws", it is not evident that there is a theory of causation embedded in the application of the experiment method to human political behavior.
If this was your position, we could engage in a discussion about the effectiveness of quantitative methods to gauge human cognition, versus the ability of qualitative methods. In other words, we can have a real discussion about methods.
I thought someone attempted to do that by bringing up fMRI, political psychology, etc. Unfortunately, I don't think you ever adressed these points.
Are political psychologists (and cognitive psychologists) who use quantitative methods less able than a qualitative political scientist to get a grasp at cognition?
I have plently of colleagues who actually know something about cognition who would be very surprised to learn that they ought to read policy papers if they are to truly understand cognition.
May I suggest you bring this up with a colleague of yours who is either a political psychologist or a cognitive psychologist? You might learn something from each other, don't you think?
"Uncovering causality requires an understanding of the fundamental problem of causal inference."
Are you referring to the problem of exogenous variables?
I say this again, the difficulty with your logic and argument is that you do not know what causes human behavior. How can anyone construct a methodology, much less apply it, without any thought as to what they are trying to measure?
RE: 9:03 a.m.
"Are political psychologists (and cognitive psychologists) who use quantitative methods less able than a qualitative political scientist to get a grasp at cognition?"
Not necessarily.
"I have plenty of colleagues who actually know something about cognition who would be very surprised to learn that they ought to read policy papers if they are to truly understand cognition."
I would assume that the political psychologist would agree that if we are trying to analyze the cognition of policyamkers, then it would be an appropriate method to examine the policy papers of these policymakers. This is where their cognition may be laid out. Do you think they would disagree?
Actually, yes I think they would disagree. I've never heard of a political psychologist whose evidence comes from reading policy papers. It just doesn't sound right to me. But perhaps you could point me to some study conducted by political psychologists that does that? I just can't think of any on the top of my head.
"Actually, yes I think they would disagree."
On what basis do you think they would disagree?
"I would assume that the political psychologist would agree that if we are trying to analyze the cognition of policyamkers, then it would be an appropriate method to examine the policy papers of these policymakers."
One what basis would you assume this?
"This is where their [policymakers] cognition may be laid out."
Now, please answer the first question.
Embedded within qualitative methods is the idea that human cognition (i.e., thinking) causes human political behavior.
So politicians who undermine their own political careers by hiring high-end prostitutes do so because they're "thinking" rather than, say, "lusting"? Seems like a rather iffy assumption to rest an entire methodological enterprise on.
"So politicians who undermine their own political careers by hiring high-end prostitutes do so because they're 'thinking' rather than, say, 'lusting'?"
Sadly, poor choices and bad judgment is part of cognition. I guess we could wish otherwise.
I would suggest the real issue is why does the media, and now political elites, consider personal matters political fodder? It used to be that Franklin and Eleanor Roosevelt had their lovers living with them in the White House. Kennedy was a well-known and public womanizer. None of it effected their political careers.
It was appropriately so. What happens in the bedroom has no bearing on political decision-making, and statecraft.
"What happens in the bedroom has no bearing on political decision-making, and statecraft."
Generally, yes. But I have two words for you:
Profumo. Affair.
"This is where their [policymakers] cognition may be laid out."
"Now, please answer the first question."
But it might all be 'cheap talk' and not at reflective of how/what the policy-maker really thinks. If you know someone will read your work, there are all kinds of Hawthorne effects to worry about.
Now, please answer the second question.
"But it might all be 'cheap talk' and not at reflective of how/what the policy-maker really thinks. If you know someone will read your work, there are all kinds of Hawthorne effects to worry about."
Maybe you are correct that a particular policy paper is just "smoke." The researcher and the reader have to determine whether a policy paper is actually reflective of the policymaker's cognition. This is an issue that real political scientists grapple with.
"This is an issue that *real* political scientists grapple with.[emph added]"
As opposed to whom? Hacks like yourself?
"Actually, yes I think they would disagree."
On what basis do you think they would disagree?
On the basis that I cannot think of a single study where political psychologists examine policy papers. That is my evidence: lack of any such study. Now, I'm happy to entertain the possibility that I'm wrong. You know more about cognition and that literature than most of us. So I'm asking again: could you please point me to a study where a political psychologist examines policy papers, a stragety that you've outlined?
If you can't, then you have no evidence, and this is consistent with my position that political psychology doesn't rely on policy papers.
So the onus is on you to disprove my statement that no such study exists (and I cannot prove the absence of a study other than by saying that I know of no such study). You can prove me wrong by pointing me to one such study.
I do not know of any self-declared political psychologists that study policy and policymaking. The closest that comes to it is James Barber in his _The Presidential Character: Predicting Performance in the White House_. (I do not know if he considered himself a political psychologist.) Unfortunately, I have not read that book -- although I may have to now. For whatever it is worth, sociologist Theda Skocpol examined policy by focusing on the cognition of policymakers -- along with other factors.
Of course, we are evading the substantive issue at hand. Would political psychologists disapprove of examining policy papers to determine cognition? I cannot imagine they would.
Actually, before we go any further...
"determine cognition"
What does this mean? Does it mean "work out how they think about things" or "describe how they think about things" or what?
"Of course, we are evading the substantive issue at hand. Would political psychologists disapprove of examining policy papers to determine cognition? I cannot imagine they would."
Wow, absolutely no data supporting your hypothesis and yet you fail to reject it anyway.
The mind of a qual on full display.
"What does this mean? Does it mean 'work out how they think about things' or 'describe how they think about things' or what?"
Both those suppositions seem fair.
"Wow, absolutely no data supporting your hypothesis and yet you fail to reject it anyway."
"The mind of a qual on full display."
I am speculating, just like you are. You are being silly and insulting for no reason. =-)
I will pose a question to you of which you have no evidence, but on which I think you have an opinion. Do you think physical scientists would object to using quantitative methods in the social sciences?
"Do you think physical scientists would object to using quantitative methods in the social sciences?"
I suspect they would be in favor.
Evidence? Let's start with...
Richardson, L.F. (1960). Statistics of deadly quarrels. Pacific Grove, CA: Boxwood Press.
"I will pose a question to you of which you have no evidence, but on which I think you have an opinion. Do you think physical scientists would object to using quantitative methods in the social sciences?"
Of course they would not object. Let me give you two examples off the top of my head. At Argonne Labs and the Santa Fe institute both traditional "hard" scientist and social scientist work together using similar methodological approaches to understand both physical and social phenomena. There are also numerous examples through history of physicist, mathematicians and economist working together.
<< Home