Defense Day

Well today is my thesis defense day. For those who are unfamiliar with the process, this is how it works at least at my university.

When you start out in a lab you do the experiments your boss tells you to do, with the goal of picking up a project. This usually involves taking up where another graduate student or post-doc left off, or reading the literature in your field and figuring out an important question to answer. Depending on how many years its been since your boss handled a pipette, he/she will suggest experiments that range from next to impossible to impossible. You spend a year screwing around, messing up experiments and hopefully generating some preliminary data to justify the research that will go into your thesis. Simultaneously you are taking graduate classes and applying to be a degree candidate. The requirements for this vary by university and department. It may require writing a review paper summarizing a related (or unrelated) field of science and defending it to a committee, or may be an actual exam in which you prove you've actually learned something in class.

Eventually, in the lab you sit down with a post doc who helps you design experiments that will actually work, and all of a sudden data arrives. If your boss asks you about the experiments he/she recommended either say you're still working on it, that you're waiting for an antibody, or pretend you don't understand English. Then, quickly distract them with the interesting data you've been generating and they should soon forget about what you haven't been doing. At this time you should be developing your possible/impossible filter, and judging your mentor's advice accordingly.

The next step is to compile all the data you've been generating into a defense of your proposal. Most labs these days have you write something that resembles a grant, because it gives you critical experience in how to get people to give you money. The importance of this skill can not be underestimated for your future success as an academic scientist. You will be justifying the financing of your science for the rest of your career, and knowing how to get funding is probably more important than knowing any specific experimental technique. You present your grant, which consists of your primary aims, methods, preliminary results, etc, again to a committee. They grill you. If your lucky your committee will be composed of a mixture of docs who will defend you from the worst iniquities of the other members and docs who won't be afraid to tell you your project actually sucks. Either way, you're in for a treat.

Once you're done with that congratulations! You are between 2 and 10 years away from graduation.

Continued....

During this time you show up in the lab for about 12 hours every day, read email and surf the internet. Between reading boingboing and arguing with trolls you run experiments and review papers. After about a year three things might occur. (1) Your idea was good, so your experiments will work and you will publish. (2) Your idea was good so you will get scooped about half way through. (3) Your idea was good, but the science gods hate you, the experiments don't work and you find out your hypothesis was totally incorrect. If 2 or 3, what do you do? Neither is the end of the world, and are incredibly common. So you come up with a new project. If you're smart, you'll figure out how to do it without defending a new proposal. Either way, you regularly meet with your committee (the one that will eventually review your thesis) and tell them what worked and what didn't. Eventually something works (or you get scooped/screwed again - start over) and you write a paper.

The fun isn't over yet. Now you have to write a thesis. This is actually fairly easy, but by now you're so sick of your project you'd rather write about anything else. So you have to suck it up. You write an introduction - basically a review of your field and why your project is relevant to it - the chapters - you cut-and-paste your papers - and a discussion. When all is said and done you basically have a book, between 80 and 500 pages long justifying why your boss has been paying your stipend for the last 2-10 years. You turn it in to your committee and sleep for about 2 days.

The last bit is a little nerve-wracking. The public defense. It's when everyone in the department shows up, along with your thesis committee and you give a talk that will hopefully show enough data that people will agree you deserve to graduate. It's not that hard, by now, no one alive should know more about the subject than you, so make sure you don't talk about anything of interest to anyone else in the department. If you do, you will be out of your element, they will ask you questions, and your 45 minute talk will take 2 hours. After that's done, everybody leaves but your committee who then spend the next couple hours grilling you about every detail of your project with the goal of inducing panic and making you think that despite all their previous enthusiasm and supportiveness, they're going to ground you at the last second. They won't, it's just hazing, but you have to accept it as the price of their service on your committee. After all, without the prospect of torturing a graduate student for a few hours there is absolutely no tangible reward for serving on a thesis committee for the last 2-10 years. Eventually they relent. If they don't, grab your belly and yell, "Ow! my baby", or if you are male say you just found out your wife/girlfriend/last one-night-stand is pregnant. Nothing is more terrifying to a thesis committee and major adviser than the prospect of a graduate student starting a family since it means they won't be in the lab as much and will constantly import colds from day care.

Congratulations! Now you're a PhD, and all your problems have just begun. You have to find a post doctoral position and write a grant to justify your existence. You will be subject to periodic reviews, more meetings, and have to travel around the world presenting your science to people who are trying to scoop you or prove you wrong. Your wife/husband likely lives on a different coast and you still make about 1/5th of what any other career requiring 25 years of education would provide. Rinse-repeat, and you've got science!

My education will delay post doc hell for a few more years as I now return to the medical wards, starting on surgery next month, and finish my training to be an MD. I then do a residency for 3-5 years. After that, if there is a Democrat in office and the country isn't bankrupt, I may then choose to pursue a research career or a mixed clinical/research career. If there is a Republican (in other words no grants, a glut of post docs and general hostility to science) you primarily practice medicine and maybe do research on the side so you don't perish from lack of publications. Last option, go to work for industry, make money, and live with that strange mixture of corporate politics, freedom from grants and moral ennui.

Anyone still want to pursue a career in science? Sounds like fun right? I'm teasing mostly, science is intellectually rewarding, you feel as if what your doing is valuable (when it's working anyway), and you're always around smart and interesting people. Further, you have a great deal of personal freedom and are occasionally rewarded by the discovery of something completely new. The problem is that many scientists feel only a small amount of their time is spent doing science, and too much of their time is administrative, writing grants, going to meetings, and otherwise trying to feed the two-headed beast of your fragile career and money-eating lab.

My advice for anyone about to start a PhD is very simple. Make sure to find a good mentor who is stable, financially and emotionally. Ideally have candidates when you are applying to grad school - but don't fixate on one doc if after working in their lab for a bit you think they are crazy or cruel. Rotate through a few labs, get to know a few mentors in the university, and learn what you like and are good at before choosing a mentor.

As far as that choice goes the major division is between old and young. The advantage of a younger mentor is that you'll get more direct interaction with your boss, learn more directly from them etc. But the risk is that they are less-established and may have to jump to another job while you're there (this has been a huge problem in the last 7 years). The advantage of a more established scientist is that they know how to survive in science, will teach you grants, and will recruit lots of talent to the lab that will be the ones to teach you how to do the actual science. The risk is that their labs tend to be big, you have to be self-motivated, and if you're not careful the boss won't notice you've been floundering for a few years. Never work in the lab of a department-head unless they have good post docs. More than anything, the people have to be right, and willing to work with you, or you will be screwed.

When you start in a lab learn at least one or two useful assays that no one else in your lab knows, and make them rock solid. You now have a hammer, look for nails. Develop an understanding of everything that is done in your lab, but don't try to master every technique. Instead barter the ones you have mastered for the ones other people have mastered. That way they are coauthors on your papers, and you get to be coauthors on theirs. Everyone is happy, the science is better, you learn to collaborate, and the division of labor ensures everyone is operating efficiently within their sphere of competence. Even if your stuff isn't working, you're still getting publications, and being a productive member of the lab.

That's all I can say for now. I think I'm going to hit the NyQuil so I can actually get some sleep tonight and be coherent tomorrow. If all goes as planned I'll be a PhD by 5PM tomorrow.