The Stimulus Is A Failure

[zakubandit]

Is there anything left to argue here? We all hate each other, you guys are so far up Obama's ass your chewing his food for him, and the other side is picking out the best spot for a new underground bunker...then everyone else who doesnt know what to do because they havent been told yet.

[DaShi]

Obama is also going to issue primae noctis, given all black men first crack at the white women.

[Drake Tungsten]

Black men already have the first crack at white women.

[KrazyHorse]

Black men already have the first crack at white women.

http://www2.gsb.columbia.edu/faculty/rfisman/RACE_resubmission_final.pdf

(page 25, table 2A)

women are significantly biased against dating outside their own race

[Drake Tungsten]

women are significantly biased against dating outside their own race

You should qualify this to only apply to white and black women. The coefficients for Hispanic and Asian women weren't statistically significant at the .05 level.

Poor Asian guys...

edit: The study seems to have been designed to measure racial preferences in potentially romantic relationships. This, of course, is different from the purely sexual relationships that I was alluding to and suspect black men might have an inherent advantage in over men of other races.

edit 2: Apparently black women at Columbia are far less picky than white women. If only I'd known!

[Ben Kenobi]

The only coefficients that are negative and individually significant are Hispanic men’s ratings of Asian women, and white and Asian women’s ratings of black men.

Also interesting.

[Drake Tungsten]

Also interesting.

Given the overall number of coefficients we estimate, however, these
three effects are likely to be spurious.

[Ben Kenobi]

You trust a liberal to analyze data correctly? I don't see why he believes these are spurious, other then his BAM.

"These are not the droids you are looking for."

Recall, however, that even though the race of the partner strongly influences
individual decisions, 47% of all matches in our data are inter-racial.

No preference would be 53 percent.

That's a pretty weak thesis.

[Drake Tungsten]

You trust a liberal to analyze data correctly?

1. Why would you assume that the authors are liberals? Two of them are business school instructors.

2. Political viewpoint has no bearing on whether someone can analyze data correctly or not.

3. No one discounts potentially interesting results in a statistical study unless there's a good reason to do so.

4. You don't seem to know anything about statistical methods.

[Ben Kenobi]

Our results indicate that even in a population of relatively progressive individuals

A natural starting point for examining the effects of background is to consider the role of the prevailing attitudes on race in subjects’ place of origin. For subjects that grew up in the United States, we use the intensity of support for a ban on inter-racial marriage in the subject’s home state

This is how they measure racism.

For subjects who did not grow up in the United States, we use the variable
RacistNeighbors_WVS, which reflects, for the subject’s country of birth, the fraction of
survey respondents in the 1990 World Values Survey (WVS) who reported that they would
not want a neighbor of another race

This is how they measure racism outside America. Why not use the same survey for Americans? Not doing so leads to errors in the analysis.

Hence, despite the fact that our subjects exhibit substantive racial preferences, they seem far more tolerant than the society at large. This is unsurprising, given the characteristics of our subject pool.
First, our subjects are highly educated, and prior survey-based research finds that same race
preferences are negatively correlated with education.

Author is willing to undercut his own thesis to make sure this gets in.

Subjects that come from intolerant places reveal stronger same race preferences. This is somewhat
surprising given the fact that our subjects are graduate students at Columbia University and
that many of them attended college away from home.

This contradicts his earlier statement and reveals the bias that education is supposed to remove this, but clearly it hasn't.

I will guarantee given my experience with liberals that these folks are very much liberals. The handwaving just confirms it for me.

[Ben Kenobi]

2. Political viewpoint has no bearing on whether someone can analyze data correctly or not.

It has everything to do with whether someone can analyze data correctly. Handwaving like this isn't justified unless they have some solid reasons to do so. Why is this data spurious? There are plenty of reasons why it could be, but their explanation doesn't tell me anything.

3. No one discounts potentially interesting results in a statistical study unless there's a good reason to do so.

Mendel's thumb anyone? People do it all the time, dismiss evidence they see as irrelevant because they find it personally distasteful.

4. You don't seem to know anything about statistical methods.

Explain to me why the data is spurious. I chuckled when I read them say that it was like, damn, this is an interesting conclusion, and then whoa, "these are not the droids you are looking for".

[Drake Tungsten]

Don't make me put you on ignore.

[Dinner]

women are significantly biased against dating outside their own race

You should qualify this to only apply to white and black women. The coefficients for Hispanic and Asian women weren't statistically significant at the .05 level.

Poor Asian guys...

I understand that the east Asian marry out rate is declining fairly rapidly these days. It used to be that the majority of east Asian women married white guys but now the majority marry east Asian guys though the marry out rate is still quite high. Also I don't like how many researchers lump all Asians together without breaking down the differences between South Asians (who have an extremely low marry out rate) with east Asians (who have fairly high marry out rates).

[Drake Tungsten]

You obviously didn't even read the study.

[Dinner]

Nope.

It's 33 pages and I'm at work.