Choosing an advisor is like choosing a spouse, and I mean "spouse" in all its myriad modern interpretations. Certainly, the consequences of your selection will be similar: The perfect advisor or spouse will forever be a source of joy in the good times and solace in the bad. But make the wrong choice and you will inevitably--like that delicious stud-muffin Meatloaf--find yourself praying for the end of time, so you can end your time with them.

As I prepared to write this column, I reread several guides to graduate school hoping to find some sage advice to pass along. I don't usually do this, but in this case it seemed worthwhile. That's because my personal history with advisors is checkered, to say the least--I burned through three in less than 2 years--and I didn't feel entirely qualified to tell anyone else how to make this decision. In fact, I felt a bit like Elizabeth Taylor (without the make-up and jewelry, of course) dispensing advice on how to find a lifetime companion. So, I turned to the "experts."

Their advice is remarkably consistent. They all agree that before you make your decision, you should talk with several potential advisors in your department, as well as to each one's current crop of students. They also suggest that you try and work on a small project with your candidate advisors. In some departments, particularly in the life sciences, this advice is formalized into "rotations," in which each grad student spends most of their first year in several different labs. By the time all this is over, both the students and the professors have a much better sense of who's going to work best where.

People in Your Neighborhood: The Permanent Postdoc

Just about every department has one permanent postdoc, and most have a whole bunch. For one reason or another, permanent postdocs have decided to forgo the tenure track while still conducting top-flight research in a university. Permanent postdocs rarely teach classes, and they typically rely on grant money to pay their salary. Regardless of their personal situation, they can be an invaluable source of inside information, particularly when you are scouting out an advisor. The good ones have likely been around the department for years, watching students and advisors come and go. So they know which advisors help their students excel and which can make students' lives a living hell.

If you're new to a place and don't know many people yet, there are a bunch of different ways to spot the permanent postdocs. You can surf over to the "People" section of the departmental Web page, where they'll likely be found under "Research Associates" or some other closely related euphemism. (Usually, this category is listed right below "Faculty" and just above "Postdocs" and "Graduate Students," a hierarchy that, by the way, also reflects their status in the department.) If that doesn't turn up any names, ask the department administrative staff. Permanent postdocs always seem to be on good terms with the administration. And if all else fails, attend a few seminars and watch where everyone sits. Permanent postdocs always choose the same seat or, as they will tell you, their seat.

Aye, there's the rub. How do you make that decision? Should you go with the guaranteed thesis from the lab of an established-but-staid professor or challenge the powers-that-be at the side the charismatic Young Turk?

The authors of these books won't tell you the answer to that question, and neither will I. I can't speak for the other authors, but the reason for my reluctance is this: I don't believe that there is just one answer. Either choice could work out swimmingly. Or not. And all the logical analysis in the world can't tell you which advisor is the best for you. Like a marriage, it all depends on the chemistry.

My first advisor-advisee relationship, for example, was a bit like an arranged marriage. I had won a fellowship to work on a thesis with someone at the Space Telescope Science Institute. The only stipulation was that I had to choose a staff scientist as my advisor. I selected a charismatic Australian theorist with an excellent scientific pedigree and similar research interests. On paper and in conversation, he seemed like a perfect match. But when it came to research, we never clicked. None of the projects we started fired my imagination and, not coincidentally, none of them got finished. In the end, we decided on an amicable split. Chalk that up as my academic divorce number one.

On the rebound, I hooked up with a Hungarian cosmologist. His work was very mathematical and computer-intensive, two of my strengths at the time, and we made a lot of progress on a small project about galaxy motions in the nearby universe. As an added bonus, I discovered that we shared a love of music. He had even released a few albums with a Hungarian rock band in the 1970s. But it was not to be. Eastern European countries weren't as open then as they are now, and visa troubles turned his annual 1-month trip to Hungary into a year-and-a-half-long hiatus. The project floundered during the separation, and we simply drifted apart. Academic divorce number two.

I temporarily gave up on research and decided to concentrate on finishing up the last of my coursework. In one of those classes, however, I came across a different sort of professor. Instead of reading out of the book, he challenged us with relevant material from more advanced texts. And he added illustrative new "real world" case studies to replace the somewhat dated examples in the text. I was intrigued and I approached him about doing a trial project. He said yes, and I had my third, and final, advisor.

Like most relationships, there were plenty of people around who told me I had made a big mistake. And in all fairness, I must admit they had a point. My advisor could be difficult to work with. (He even admitted to occasionally antagonizing fellow faculty members in hopes of avoiding committee work.) But somehow the combination worked and we started cranking out papers. Even after I defended my thesis and moved to a new university to take a postdoc, we started a new project that bore fruit right up until the time I left academic research.

So, it took me three academic marriages to find the right advisor. What does that mean for you?

It means that in selecting advisors (or spouses), you have to follow your heart. By all means do your research, find some good candidates, and use logic to rule out the obviously bad ones. But when the time comes to choose, trust your instincts.

Now you have a school and an advisor. There is only one thing left to do--find a thesis topic! Tune in next month as we take an inside look at the formative stages of a doctoral thesis.