Showing posts with label research. Show all posts
Showing posts with label research. Show all posts

Tuesday, May 27, 2008

Working Papers at CERGE-EI

There is a plenty of rules for submission and evaluation of Working Papers at CERGE-EI. Most of them attempt to smooth the process and limit its length. It is all governed by one person I will not name, because this person changes and for another reason that will be obvious in a minute.
This person receives the paper, selects the referee from CERGE-EI faculty, usually based on suggestions provided by the author, sends them the paper, collects the reviews etc. The referees are supposed to be done in 3 weeks. There are no limits on how fast they need to be chosen. So check this:

This is a submission email:

Date: Wed, 23 Apr 2008 13:33:33 -0700

And this is an email send by that person to the referees:
Tue, 06 May 2008 12:51:54 +0200

Yes, it took him 2 weeks to send a 3 line email. Given our history, I guess it is personal...

Monday, May 26, 2008

Emptiness

I sent my third paper to my dissertation committee and I'm await comments. Even though I'm probably still far away from finishing the dissertation, I experience a strange feeling of emptiness. I surely have what to do (some magazines and fiction books to read), or I might even work, but I feel strange anyway. I actually did start looking for fourth topic, but I was google old idea solved in a few seconds. But I also don't feel like I need to - I'm not looking for a job at a university, and I don't even plan to continue the research. It's probably time to polish existing texts.
P.S. I'm required to have three papers to finish my dissertation. Thus, I'm quite close, unless my committee rejects the paper.
After two weeks of waiting, first comments came. They are positive. Two more to go...

Wednesday, April 23, 2008

Czech Academy of Sciences

... is famous again. But not for what you would expect:

But the IFPI says that the research institute was in fact the unwitting host of a server that powered one of the largest pre-release music archives in the world, and with the help of Czech police, the server and its 4TB of data has been shut down.


Given my insider knowledge into piracy at Czech Universities, I'm not sure that "unwitting" is the right word. Still, 4Tb of pre-release music? Even at 700MB/CD (ie, no compression whatsoever), it is about 5700 CDs. There is that much pre-release music?

Monday, March 31, 2008

Predicting future

Predicting future is hard, even in the short term. Yet, it is fun to read about old predictions. This one is 40 years old and quite interesting.
It would be great to see how the world will look like in 400 years, but I hope 40 years or so will have to be enough. For this generation

Runaway health care costs

Health care system in the US is extremely expensive. Yet, it seems that quality does not necessarily reflect these costs. For example, it is sometimes argued that incentives are poorly set: doctors do unnecessary test of *insured) patients, who don't have many reasons to care about the costs.
Yet it seems at least possible if not plausible, that the US bears disproportionate share of research costs that lead to (significantly) higher prices of new treatments. I'm not sure how significant this can be, but maybe European systems are cheaper also because these system free-ride. Well, other reason can be that doctors cannot afford a Porsche in Europe (at least in Czech Republic).

Sunday, March 16, 2008

Sanity Check

Writing a dissertation is difficult for many reasons. Primarily, it is very independent, very long-term work with uncertain outcome. It is very hard to evaluate progress or lack thereof. And even if you notice lack of the progress, it is hard to really know why.
You may be doing progress faster then usual yet fail to notice, you may be working on the wrong idea/topic, or you may be simply too stupid to produce anything original and useful.
I wasn't happy with my progress from the very beginning and it was driving me crazy for a long time. I'm still not happy, but finally I found a way how to keep myself sane. Here it how it works. Remind yourself that

1. You are (among) the best. Yes, you are - no matter how do you feel about it, you are easily in top 1% of the world population in terms of education, intellectual ability, knowledge and/or future income.

2. You can do it. Seriously, most people finish and you probably know plenty of those whose research is appalling by your standards. No matter what you produce it will be better than that, which is more than enough.

3. End is close (this helps if you are at least in the middle). It does not seems so right now, but you can probably think of steps you need to take to finish. Split the task into small pieces, make sure that the timing is reasonable and get to work. It will be over sooner than you think. Remember, the time is running faster in the second half.

4. Quality matters, but you need to choose the right scale. You have read hundreds of papers, most of them published in the best journals or written by the "gurus" of your niche. Don't expect that your first paper will be as good or produced as fast as papers written with 20 years of experience. Strive for quality, make sure that your work makes basic sense, but remember that most than 1/2 of the publish research is never referred to.

5. You are living the worst and the best time of your life. On one hand, you are totally independent. If you take a day off, most likely (depends on your field) nobody notices. Even if you take a week of, it would be OK. On the other hand, you are your own boss and you can be pretty tough, can't you? Also, you don't have any money. Don't worry, there is probably plenty of money waiting for you in the future, but you will have no time to enjoy them.

6. It is enough to finish just this part and the rest will be easy. Choose a specific goal, finish it and then choose another one. It will be getting easier and easier. First paper is the most difficult.


7. Find a co-author, ideally not your adviser. Dissertation committees don't like papers co-authored by your adviser, but other students as co-authors are welcomed. Find one for at least one paper, possibly (depends on your school and your job market goals) two. If you get stuck, they will help you to move along and produce something. If they are more experienced than you, even better. You will learn by doing and finish a part of your dissertation along the way.

8. Visit presentation by other graduate students. It helps to keep you informed how they are doing and you never know when you meet even bigger looser than yourself and thus improve your self-esteem.

Thing I did not do but that can help a lot.

9. Finish first paper as soon as possible. Nothing helps more than to know that 1/3 is over. Seriously. If you have 3 papers in various stages of progress, but none of them published, it feels like you have nothing! (I know, I live it every day). Focus on publishing one paper, if it is possible (and makes sense).


Repeat the previous points to yourself as often as needed, focus on those that help you most. And relax, it's fun. Or perhaps, it will look like fun when you will be 60 years old.

Sunday, January 27, 2008

Slow progress and multitasking

At the beginning of my PhD, and especially after I clarified my research ideas, I was hoping for fast progress. As usual, it did not realize. I had a couple of alternative explanations (topic is bad, research is difficult, I'm not smart enough, I'm not working hard enough).
After some thinking, I found out that the problem is concentration and multitasking. I'm not talking about working on one idea in the morning and on the other in the afternoon (and writing a blog in the evening), but about solving a model and say, talking to your friends over IM and reading a newspaper. The multitasking killed my ability to focus, to remember and to work hard.
There is a research why, but I'm still looking for how to reverse the process.

Monday, January 14, 2008

Real world...

Steven Levitt is a big star. Among his famous articles are those analyzing gang members selling drugs and cheating among Sumo wrestlers. See his list of papers for much more.
One would expect that you cannot get more astounding. Well, you would be wrong. His latest paper analyzes prostitution in Chicago. Results are, as usual, surprising. Condoms are rarely used (25%), yearly wage is about $20,000 (I guess they should have a union to reduce supply) and pimps are, in fact, a good thing for the prostitutes. Police seems to look away if you give them a freebie. But the craziest part is at the very end. I have to quote, because this is just unbelievable.
Last client's report "revenue $200, father & son; inside client's house while mother watched and masturbated".
Well, Levitt may be a superstar, but I have serious doubts about his data. Or about this world.

Thursday, November 29, 2007

The next small thing

Many people are trying to predict the "next big thing" - the invention that will change our world like computers, internet or mobile phones did. The successful ones will undoubtedly make big money on it.
But I have notice many small "big things" that are less known, probably less important yet very helpful, too. For example LED lights and reflective materials.LED lights are interesting because they have very large expected lifetime (around 100,000 hours) and they are very effective (no heat is generated). Notably, most trucks in the US use them already; they are also used in the street lights everywhere. The reflective materials are used on many things - street signs for better visibility in night, clothes etc. They are able to reflect the very little light that hits them back to the source (I'm not sure I understand how they work in detail), which makes them VERY visible in night.
The combination of these two is cool, too. I was riding a bike with a LED light that lasts more than 20 hours with three small AAA batteries, while generating so much light. And all street signs and warnings were so visible I did not have to be afraid of riding of the road etc.
These small things are pretty much trivial, right? Yet they improved our lives significantly for better, in certain areas.
This makes me wonder what will come next. Big hopes are on the carbon nanotubes, but something else might come first. For example, the mechanism by which certain plants (like water lilies) keep themselves clean using only water seems to be understood, allowing it to be replicated in factories. The secret lies in a special structure of their surface that makes cleaning very simple (using water and no pressure). Image the possibilities - easy to clean kitchen, shoes, cars, walls that can be cleaned from the dust or graffiti with just water hose.
This or any other small thing will not change the way we live (LEDs did not either), but it can make life much easier.