I moved!

To my very own domain.

Tenure and Other Concerns About The Academy

Reading this post on Steven Levitt's blog got me thinking about the issue of tenure (thanks for the link, Zack). The main argument Levitt makes is that tenure does not increase overall productivity in academia because it is too much of an incentive early in the careers of academics and not enough of one later on. On the other hand, the reason we have tenure is that research can be risky and the benefits can be great if the research is successful. Furthermore, at stake is academic freedom: no political climate should be able to stop academics from expressing their thoughts. Both of these points are argued in the comments of the aforementioned post.

Mind you what you are about to read are the perspectives of an inexperienced grad student who has been in contact with academia only a few years, so read on with a grain of salt.

I think the tenure system is a bit broken. Academics don't get too much academic freedom that they wouldn't otherwise have. They don't pursue risky research early in their careers in order to obtain results that lead to the golden dream that is tenure and can become rather unproductive afterward. Of course this doesn't apply to every academic out there, but I can certainly think of at least a few cases. So, it is likely that a university gets stuck with an unproductive employee who can not be fired. While I understand where the idea of tenure comes from, it seems to me that it needs some tweaking.

Why do we think that the same person who is good at doing research should be good at teaching and at administering an academic department? Why do we think that this person should have infinite job security almost irrespective of his performance? How about those not being the same person? The only people that need to have some kind of risk-insurance in a department are those who do research.

I think that there need to be different contracts for different kinds of positions within an academic department. That is, hire people whose main responsibility is either to teach, research or administrate -- not all three. Then the processes for evaluating performance can then be tailored to each kind of contract.

For those who are hired as teachers, their performance is evaluated once every few years based on student class reviews, competence in the subjects they are required to teach and so forth. For those hired as researchers the performance evaluation can be tailored to the research they will be pursuing. "Tenure time" could be handled in ways similar to those used to allocate research funds: via proposals and committees. That is, a committee decides how much time of infinite job-security a researcher gets. During that time the researcher needs to produce results and a her next proposal and then the cycle repeats. This is where the people administrating the department come in. They are the ones in charge of hiring people, evaluating the performance of the department and adjusting accordingly. Perhaps there could even be mechanisms set up to switch roles within a department.

The first problem with some arrangement of this sort I can foresee is that there seem to be very few incentives to become a mere administrator or a teacher. All in all, most of us are in the game for the glory of doing research (maybe not?...), not for the red tape or those pesky undergrads asking questions about basic subjects during office hours! However, hopefully financial and other kinds of incentives could be used to make up the difference.

Another important caveat is that tenure offers a convenient facet when it comes to hiring-committees. They can hire the best possible candidate without fear of hiring their replacement. If there is a mechanism for switching from an administrative contract to a research one, why would an administrator planning on trying to become a researcher hire the competition? I have no good answer to this one, other than perhaps there should not be mechanisms for switching between administrative and research positions. This does make the administrative positions seem very unappealing...

Yet another caveat: how is the performance of the administrators evaluated? Maybe restructuring academic departments in this fashion has implications that require a broader restructuring of the university bureaucracy.

Yet another important caveat is the one about making the transition from the current system to whatever new system. I imagine a university changing their tenure structure to have less risk-insurance would have a harder time hiring people. On the other hand, it seems that the market of newly graduated PhDs and postdocs has enough of a labor surplus that universities could get away with it.

What do you think?

Top 10 Pieces of Advice I Wish Someone Had Given Me When I Was an Undergrad

Like with any advice, the best advice I can give you about it is that you take what you feel applies to you and toss the rest. In any event, here it is...

10. Explore and get to know yourself, then establish your long-term goals as early as possible. Establish intermediate goals to get you there. Formulate the immediate plan. Execute it. Revise your goals. Redo the last three steps over and over.

9. Lead a healthy lifestyle both physically and mentally. Eat good food, exercise, go outside. Broaden your horizons. Meet people. Learn things unrelated to your field.

8. Get rid of your television.

7. You are too young to be bored. If you are bored it either means you are not pushing yourself enough or not living enough (or both). Don't let yourself be bored. Keep busy. Push yourself.

6. Make the most out of your summers. 15-week vacations will be a lot less frequent later on. A summer without learning or experiencing something significant is a wasted one.

5. Have an independent project in the back or front burner at all times. Have a book you are reading in your off-time (e.g. while riding the bus). Podcasts and books on tape make for great workout music. Finish one such book every few weeks.

4. Read the textbook from cover to cover. Do every problem in the book even if that is not required. It is a pity to pay over $100 for a book you never read. Well... there are some books that are just not worth the effort, but at least do this with the ones you know are important.

3. Be dedicated but make sure you live. You only get to turn 21 once. Make sure your friends tell you what happened. Never spend more than 3 weekends in a row without going to a social gathering and having a great time. Get hammered.

2. Travel.

1. Take your education into your own hands. Don't expect that you will learn by just following professor's, university and program prescriptions. Pursue your interests on your own in as much depth as time allows. Then pursue them some more even at the expense of some schoolwork.

The Hidden Iraq

Just watched yet another documentary on Iraq. The most astounding fact revealed by it was that there have been 93 journalists killed during the conflict. To give perspective, about (sources vary) 70 journalists died during Vietnam (1955-75). The refreshing bit is that the internet seems to be having an impact on this. The documentary finishes by talking about bloggers from Iraq whom by virtue of fitting in are not targeted like western journalists and manage to get first-hand accounts out via the internet. This makes me wonder about the status of internet availability, though. And even if it is available, I wonder if there is government/US control of it and how much. Furthermore, it seems that it is a matter of time until whomever is targeting western journalists starts targeting the bloggers.

A Failed Coup...

Just stumbled upon this and watched instead of doing my quantum homework. It is about events that took place in Venezuela 2002. I'm largely ignorant about all the subtleties of the matter, but I can still say that the documentary is rather insightful in that the makers seem to be very close to Chavez himself during everything. This same feature makes it a bit biased perhaps. One way or another, worth the watch.

Open Source Knowledge

The lack of posts is Jackson's fault.

The overly inactive open-source revolutionary in me spoke the words 'if you ever write a textbook, please make it free' to two different professors this week. Should books be free like Linux is free? Yes, they should. Of course, printing costs are a reasonable thing to charge for and the author's time also has monetary value (which open-source software authors are usually willing to forfeit as probably would some academics). As it is there exist plenty free 'text books' online. Don't believe me? Check this out. Online books are nice, but if you are going to be doing a fair amount of reading it is actually nice to have a reasonable quality hard copy. So, my rough sketch of a solution is to create a 'publishing' company that operates as follows.

Books are made available online for free or for minimal cost (to break even in terms of hosting). People can then order printed copies of the books for an additional cost to cover printing expenses. This is all organized through a web-site... duh, it is me coming up with it, right?

This has its logistical fallacies: it is probably a lot cheaper to print several thousand copies of a book than printing them 'on demand'. But with reasonable popularity measures and printing practices the logistic problems could perhaps be made into manageable ones.
EDIT: Doing about 10 minutes worth of research into publishing costs revealed to me that one can expect to get a paperback quality book published for $5-20 per copy depending on the number of copies. This definitely makes the idea plausible.

Now we can take this to a whole new level. Imagine web-book 2.0. Users can register and have reviews, rankings and recommendations for books. Perhaps one can allow users to upload materials and have the system keep track of what items are related in one way or another. As the number of books and users grow the costs would probably do so as well, but I think that a non-for-profit spirit plus some creativity in ways to earn revenue could make ends meet. For instance, printed books could be sold at a price slightly higher than cost as to cover the other expenses of the site. Or perhaps users could pay a yearly membership fee.

Now a bit of due credit. Zack had a variation of this idea several years ago and it has been in the back of my head since then.

The main problem with all this is that for whatever reason the books that are available for free online are rarely as good as the overpriced published ones. I can think of several possible reasons why this is so, but whatever. In any event, my generation might have just gotten fed up enough with new editions that merely shuffle problems around and price tags over $100 to start producing reasonably good content that can be made available in a non-profit manner. I say it is the duty of our generation to start by writing a better e&m book.

Viva la Linux revolucion!

Security Profiteers

I've been meaning to blog about this since my Florida trip in December. I had a long wait at the Orlando airport and while I was wondering around I stumbled upon a little light-blue and white kiosk that looked as if it came straight out of apple. The name: clear. There was nobody there, but there were some pamphlets. The deal is that you pay this company $100 per year and you get to skip airport security. Bruce Schneier has already blogged about this one.

The current security measures at airports are ineffective at best. They do more to hassle passengers than prevent terrorist attacks and they are a waste of resources. The main problem with something like Clear is that it seems to be out there just to make profits from having a security checkpoint that is run the way they all should be.

Graduate Studies: Some Ideas

Francesca was telling me about how they do classes at her undergrad school in Italy and a few things seem to be very good. After some more thought, this post came about.

Three things:
There are exceptions but in my experience I've usually felt like professors would rather be spending their time in ways other than teaching.

Secondly, I do most of my learning by reading the prescribed and other books. I think I'm not quite the norm here, but it seems that there is enough literature out there that one can do most learning from books. Of course, there is still trickery and guidance that must be passed on from generation to generation in more interactive ways and books certainly don't answer questions -- not even if you raise your voice, curse, pull your hair or ask nicely!

Finally, it seems safe to assume that graduate students have enough interest and motivation to pursue their own studies without much of someone forcing them to do things like turning in homework assignments for grade in order to make sure they practice problems. Besides, the qualifier will speak for itself.

With all this in mind, it seems that the concept of 'graduate course' could use a bit of revamping. The main idea is to give more freedom to the students while trying to minimize the time that faculty have to devote to them in terms of the basic subjects -- research is a whole different beast. A department could, for instance, run the program so that the students are not required to take a core curriculum of classes. Instead, simply make clear what subjects should the students be responsible for. Learn this. See ya at the qual.

Of course only that would render the department useless. They need to do more for the student, and they may also want to monitor progress. One minimum requirement is for the department to provide tools such as problems and solutions (most schools give old qualifiers to current students as it is) -- according to Francesca in Italy this is fairly standard practice in most classes. Another tool to be provided should be an 'official bibliography' which should include information about what chapters to study. Heaven forbid someone read all of Jackson in hopes of preparing for the qual! Each student should be assigned an academic adviser with whom she is required to meet a few times a term. Finally, instead of the department assigning someone to teach each subject, the department could simply designate a faculty member to be the 'resident expert' in each subject for the year. The 'resident expert' should be available to students a few hours a week and could perhaps host a 2-3 hour group Q&A session each week. One could request that questions be emailed a couple of days in advance so that the Q&A sessions run a bit more smoothly.

Such an approach could have many benefits. For one, research is largely a self-study type activity and is also rather unstructured. I've heard people say that making the transition from classes to research is difficult because one has been trained in a very structured fashion for many years. For another, I think placing most of the responsibility about what and how to learn on the student is valuable preparation. Again, we are trying to train people to be capable of learning things that nobody knows yet. Finally, I think it would allow departments to use their faculty resources better. For example, I've asked a few people why is it that physics departments usually don't teach an undergraduate freshman class on basic mathematical methods for physicists. The answer is invariably: "we would like to, but it is hard to find resources for such a thing." It drains a lot less time to be available by appointment than to prepare lecture, hold office hours and grade homeworks and tests i.e. being the 'resident expert' should be a lot less time consuming than teaching the class.

As far as monitoring the progress of students, the department could require that the student turn in problems a few times a term. Checking for satisfactory progress should be a bit quicker and easier than assigning grades fairly based on homework and tests. In the end, the ultimate progress indicator is the qualifier anyways. But of course, departments may want to ensure that nobody is getting a one-year paid vacation!

As I write this I am thinking of Physics programs, but I have just edited the title to take Physics out. This, with some variation, could apply in most subjects.

EDIT:

A re-formatting of this kind would obviously be very difficult for any department as there are probably a million bureaucratic issues to deal with. More importantly, the current system clearly works (maybe it could be better, but it is certainly good enough). So perhaps a dose of "if it ain't broke don't fix it" is in order. There are also several advantages of the current system. My friend David pointed out to me that a class is a good way to set pace and to get people with similar interest focused on the same subject. Also there is that a lecture by a good teacher who likes to teach is likely worth more than the same time spent reading on the same subject.

So maybe all this should just be some possible suggestions on how to run classes rather than how to run the first year of a graduate program.

Income Inequality and Human Nature

An interesting read from yahoo finance. Most of the arguments should seem familiar: gini indices and crime rates. However, this bit was interesting and a bit surprising:

There's a very interesting strain of economic research showing that our sense of well-being is determined more by our relative wealth than by our absolute wealth.

In other words, we care less about how much money we have than we do about how much money we have relative to everyone else. In a fascinating survey, Cornell economist Robert Frank found that a majority of Americans would prefer to earn $100,000 while everyone else earns $85,000, rather than earning $110,000 while everyone else earns $200,000.

Think about it: People would prefer to have less stuff, as long as they have more stuff than the neighbors.

I suppose it is important to remember that economics is greatly impacted by psychology.

Things To Do With $100 Billion

The following thoughts originated while discussing a completely different matter with Prof. Hsu at his blog.
As per census.gov there are about 1.2 million k-12 teachers in the US. We could give each of them a $50k salary increase and have $40 billion left over to provide whatever other needs schools have which would yield about $400k for every school (I estimate there are ~100k schools in the US).

Alternatively, we could give every one of the 7.1 million college students who receive financial aid a $14k break on their loans.

We could also do something for people who can't afford health insurance, or treat war veterans better, or many other decent things. Unfortunately, our priorities are not such. Granted the $100bn price tag is payable over the course of a few years, but any of the aforementioned causes would benefit greatly from the money disbursed over that period of time.