wannalearnecon

I found a useful link for an economics NSF essay that won the fellowship: JulianReif.com

Does anybody know if there are sample NSF essays posted online anywhere? It would be nice to see what kind of quality is necessary to win an NSF. Also, if anybody knows of a site that has good tips on the NSF, that would be cool too...

Yes, Octave is a pretty good substitute. I've used it myself.

I can confirm Elly's comment. I received an award letter too. :)

From my conversations with older economists, that article seems pretty much right on.

A masters degree (in anything) offers little to no advantage at most litigation/econ consulting firms. You would most likely be starting at the same title/pay level as BA's. The only things you can bring to the table that undergrads don't have is 1) PhD 2) worthwhile experience working somewhere

try this page: Normal Distribution -- from Wolfram MathWorld

When I took the test in 2005 I definitely thought that the quant section was MUCH harder than the practice tests I had taken, including ETS's. I also ran out of time because it was so unexpected for me. I think if you retake it and get 760+ most adcoms will view the 600 as an aberration though (especially if you include a note mentioning this).

Outside of Harvard and MIT, my understanding is that these numbers are fairly typical of top programs. Chicago tends to fail more people early on in the program while other schools tend to "not graduate" more in the later parts of the program.

According to official information: 1. On average, 65% pass all three core exams on their first try 2. On average, more than 80% pass all three core exams after two tries 3. On average, more than 90% who pass the core are eventually granted a Ph.D. You only get two tries to pass the core, which is offered only once a year.

I am currently in graduate school and have an opportunity to be published in a peer-reviewed journal ranked in the bottom third of the econphd.net journal rankings. The paper relates to research I did prior to graduate school. Given that it would take me little-to-no additional effort to have this paper published, is there any reason not to do it? I am worried that because the journal is not top-level it might hurt my career (my assumption is that my research in graduate school will be worthy of publication in a top journal).

I thought applicants had to apply within the first 12 months of their graduate study, meaning you could only apply in your first year of grad school. Did I miss something?

Definitely take a look at work done by Perisco, a professor at NYU. He has done some interesting research using outcomes-based testing (as opposed to regressions).

Ah yes, "continuously" differentiable would be the key. As usual, my inability to notice important key words was the problem. Thanks guys!

I am using the following website to learn some real analysis: http://web01.shu.edu/projects/reals/cont/index.html I am confused by the C(n) function in chapter 6. It is a function which is n-times differentiable, but not (n+1) times. The function is: g(x) = cube root of x^(3n+1) if n=1, g(x) = x^(4/3). But this is differentiable more than once (g''(x) = (4/9)x^(-2/3) ), so I don't understand. What did I miss here?

http://web01.shu.edu/projects/reals/reals.html

Yes I've used SVN in a work environment before and liked it but the setup (which I've never done before) looks a bit daunting. Is CVS or RCS easier to use?

Does anybody here have suggestions for good version control software to use? (i.e., software that keeps track of different drafts of papers).

Quick Real Analysis question: How do I prove that the product of 2 continuous functions is continuous? I am stuck here: Need to prove |f(x)g(x)-f©g©| Assume we find sufficient delta, |x-c| =|f(x)g(x) + f(x)g©-f(x)g© - f©g©| What is the next step? Maybe it's just late, but I feel like I'm missing something obvious. Thanks!