Polerius -
According to your cite, 87% of the respondents wanted the “under God” clause retained. Only 12% wanted it removed.
I don’t think you can get from there anywhere near “Democrats are more likely to have a college degree than Republicans”.
Regards,
Shodan
A particular aphorism seems appropriate here. (paraphrasing winston churchill, at least supposedly.)
A young man who isn’t a liberal is heartless.
An old man who isn’t a conservative… is brainless.
Don’t necessarily think it’s entirely true, but maybe not entirely false either.
Yet more wisdom from non-pompous, non-arrogant, non-holier-than-thou conservatives :rolleyes:
Read my earlier post. From:
(1) College graduates were more likely than those who did not have a
college degree to say the phrase “under God” should be removed.
(2) Democrats and independents were more likely than Republicans
to think the phrase should be taken out
we get:
The percentage of Democrats who are college graduates is
higher than the percentage of Republicans who are college graduates.
The percentage difference may be small, but the above follows from (1) and (2).
No, I am afraid it does not.
For all we know, there could have been 119 independents with college degrees who wanted the phrase removed - and one high-school drop out Democrat. Your cite doesn’t say.
It would be a little like lumping house cats and tigers together, comparing them to schnauzers, and saying “tigers and other cats are more likely to eat humans than dogs”. Add to that the fact that wanting to remove the “under God” phrase is not as good a proxy for “being a Democrat” as actually voting Democrat, and you may see the problems.
My CNN cite indicates that college degree holders split Bush-Kerry roughly fifty-fifty. The two groups that went for Kerry were the most educated, and the least. An interesting demographic, but very problematic for those who want to argue that liberals are smarter. Since, it seems the same message appealed both to the under- and the over-educated. 
Regards,
Shodan
Short answer: The general literature on how college affects students.
Long answer: Will have to wait until tomorrow. My Pascarella & Terenzini is at the office. They’re the guys who pretty much summarized every good study ever written about How College Affects Students–a big enough book I don’t exactly haul it back and forth in my attache case.
I don’t know if it’s the bull sessions or what. That’s an intriguing question, really, because I’m with you in thinking it ought to be the case that the sage, experienced person in the front of the lecture hall is the one who sways student thinking.
I believe research has shown that students don’t, in general, change their political perspectives a lot in college–but they may move slightly towards the overall bent of their college peers (or the select peer group they seek out).
You are right. The correct conclusion from (1) and (2) is:
The percentage of Democrats and Independents who are college graduates is higher than the percentage of Republicans who are college graduates.
I was assuming that Independents were a negligible percentage compared to Republicans and Democrats, but in fact, according to your cite, Independents are 26% (vs 37% for Democrats and 37% for Republicans)
As you mention, your cite doesn’t tell us what percentage of Republicans and Democrats are college graduates. It simply tells us about what percentage of Bush voters and Kerry voters are college graduates.
So, the question of what percentage of Republicans and Democrats are college graduates has to be settled using other data (not that it is an important issue, but is was something that the OP asked about)
From The Pew Research Center - The Republicans, who are they?
Total Total
Total Republican Democrat Independent Republican Democrat
% % % % % %
College Grad. 21 25 19 20 24 20
Some College 23 24 21 23 24 21
High School Grad. 38 38 39 38 38 39
Less than H.S. Grad 18 13 21 18 14 20
The biggest gap between the parties is among the least educated, with the Democrats holding a 6 point lead. The second biggest gap is among college graduates, with Republicans holding a 4 point lead.
Interesting. If true, it would argue in favor of the theory that the disproportion of liberals in post-graduate programs and professorship is due more to selection of the already liberal rather than bringing moderates and conservatives over to the dark side. 
Regards,
Shodan
Yokay, as promised:
Generally speaking, college has a “liberalizing” effect but it appears to be less today that it was, say, back in the 1960s. Most of the change is seen among those students who enter college “middle of the road.” Conservatives stay conservative, liberals stay liberal (and don’t generally become more radical).
Living in residence halls seem to be more liberalizing than living off campus or living in greek housing.
They sum up the section thusly: “These findings and others suggest that the influence of peers may be as great or greater than that of faculty . . . the proportion of peers who hold liberal (or conservative) attitudes has been shown to be a source of influence on shifts in political liberalism.”
There’s been no small amount of research on this political stuff, but educators tend to be more interested in other sorts of values that change (nor not!) during college.
Incidentally, they’ve updated this book; I need to get a new one because this wouldn’t cover work done in the 90s.
Why would it have any bearing on whether the campus was liberal? Speaking from someone who was in the Greek system at large public university, I found no significant difference in the political beliefs of those in fraternities and sororities than the undergraduate student population as a whole. And yes, there has been research that confirms that observation. Time to hang up the stereotypes.
I find it disturbing that ANY text book for teaching writing skills has 12.5% of ANY political slant.
I would want to be taught writing, without getting a political side-lecture thrown in.
Well, the writing examples have to be about something, and the fact is that many of the world’s most exemplary writers have spent their time writing on issues of great political moment.
Great writers also write about people, about the human condition, and it’s very difficult to do this without taking a position on certain issues.
No, it does not.
It’s possible, but it’s not logically necessary. Lord knows how you came up with this.
It’s very easy to prove incorrect, even without having indepedents in the mix. Suppose you have 1,000 college graduates, 550 Republican and 450 Democrat, and 1,000 non-graduates, 450 Republican and 550 Democrat. Of those, 450 Democrat college grads and 100 Republican college grads want “under God” removed from the PofA, and 200 Democrat non-grads and 0 Republican non-grads want it to stay.
There; I’ve now created a perfectly plausible scenario in which
(1) College graduates (550 out of 1000) are more likely than non-college graduates (200 out of 1000) to want the phrase removed,
(2) Democrats are more likely ( 650 to 100) than Republicans to want it removed, BUT
(3) Your conclusion is still false. The Republicans have more college grads, 550 to 450.
Very interesting. I’ll have to think about why the above works out the way it does, but for now, I’ll show the method I used to come to the conclusion, to see what went wrong in my proof
First, some definitions:
g_r = fraction of Reps that want "under God" removed from the Pledge
g_d = fraction of Dems that want "under God" removed from the Pledge
g_c = fraction of college grads that want "under God" removed from the Pledge
g_n = fraction of non college grads that want "under God" removed from Pledge
c_r = fraction of Reps who are college grads
c_d = fraction of Dems who are college grads
Statements (1) and (2) above can be written as
(1) g_c > g_n
(2) g_d > g_r
From above definitions, we have
g_r = c_r*g_c + (1-c_r)*g_n
g_d = c_d*g_c + (1-c_d)*g_n
which can be re-written as
g_r = c_r*(g_c-g_n) + g_n
g_d = c_d*(g_c-g_n) + g_n
From (2) we have g_d > g_r, which means
c_d*(g_c-g_n) + g_n > c_r*(g_c-g_n) + g_n
==>
c_d*(g_c-g_n) > c_r*(g_c-g_n)
Now, from (1) we have g_c > g_n, which means
g_c - g_n > 0
which implies that the previous inequality reduces to
c_d > c_r
i.e. the fraction of Dems who are college grads
is larger than the framction of Reps who are college grads.
which is the conclusion I mentioned. Either something is wrong in the above derivation, or something is wrong in your example.
In the end, of course, it doesn’t matter, because even if the conclusion were correct without the independents, if we include the independents, the conclusion no longer holds. And we have direct data from Sam Stone, so we don’t really need this indirect method of getting at this information.
Nevertheless, I’m curious as to what’s wrong in the derivation above.
OK, I think the mistake was in these two equations:
g_r = c_r*g_c + (1-c_r)*g_n
g_d = c_d*g_c + (1-c_d)*g_n
They shoud have been:
g_r = c_r*g_{c|r} + (1-c_r)*g_{n|r}
g_d = c_d*g_{c|d} + (1-c_d)*g_{n|d}
where
g_{c|r} = fraction of college grads that want "under God" removed from the
Pledge, given that they are Republican
g_{c|d} = fraction of college grads that want "under God" removed from the
Pledge, given that they are Democrat