Wednesday, March 22, 2006

Shorter Postdocs Would Be Better for Scientific Progress

I've been going to a lot of seminars lately. Since the trend of scientific training these days is:

grad school 5-7 years
postdoc 5-9 years

I've noticed this has a severe effect on the way we think.

We need new blood.

For a while, there was a lot of emphasis on making sure that graduate students switched fields for postdoc, to 'get exposure' and 'broaden their training.'

In reality, these people take longer to get jobs, because they either

a) have to spend the time to get a footing in the new field
b) have to spend time at the end of their postdoc going back to their original field.

So here's a thought. In the 'old days', postdoc 'training' was only 1-3 years long. This had some interesting consequences.

First, let's think about how labs are structured.

1.You have the lifers: lab managers, technicians, and PIs.
2. You have the long-termers: grad students.
3. You have the fast-moving component: postdocs.

Oops, except now the last two catgeories have kind of blended into one long, slogging pile of people who are in no hurry to go anywhere, and who get entrenched in thinking, more and more like the lifers.

I've noticed that many advisors thrive on having new people in the lab: it's like a new toy. Everyone I've worked for was excited about me at the beginning, but after a while they lose interest. And it's not just me, I've seen this happen to everyone.

So here's an idea. Maybe if we went back to shorter postdocs, it would help invigorate science more. Speed up the mixing process and encourage more cross-discipline collaboration.

Let's just leave the job thing out of it for a minute.

Okay, minute's over. I still have no solution for that.

Labels: ,


At 11:29 AM, Blogger Matthew said...

Frankly, I don't understand the whole system. If I have a Ph.D, I damn well better be able to get a job that requires a Ph.D. Requiring "postdoc" for a scientist is sucking more life out of the postdocs for less money than they should be getting.

At 1:06 PM, Blogger coturnix said...

I am seriously considering doing a post-doc in my field, in a neighborhood lab. It will cut the time short, make my research output more focused as well as of a higher quality (lots of unpublished data to build on). They used to do post-docs with their graduate advisors back in the 1960s-70s. Why not again?

At 1:33 PM, Anonymous Anonymous said...

I agree. I switched fields--both organism and approach, and it's taken me a long time to get a project up and running. I was encouraged to make the changes that I did, because it would make me more well-rounded and help with funding. Which it did. But there's no penalty to staying in the same field--the folks that ignored that advice and are doing their postdoc in the same field are getting funding and getting jobs, and they are doing much shorter postdocs.

At 2:17 PM, Blogger dlamming said...

I disagree with a lot of your observations here. Maybe it field-specific, and of course some of it is lab specific, but anyway, here goes...

Generally, grad students are the long termers. Most labs, assuming they have techs and lab managers, will go through 2-3 techs over the course of a grad students career (which, yes, is 5-7 years). There are a few exceptions to this, but most good lab techs/managers head off either to grad school or industry after a few years, and the bad ones leave due to friction with the boss.

The component with the fastest turnover then is the techs. The longest is the grad students.

Postdocs? Generally, in the middle. I grant you that there are 10-year postdocs, but most people I know or have seen either do a short postdoc (~2 years) and don't stay in academia, or go 3-5 years and produce some good work. Now, some (many?) of these go on to do a second postdoc, but it's always in another lab. Often, this is for family reasons, but obviously you're going to be more successful stirring the pot and getting a fresh perspective on things.

With one possible exception, I've never met a postdoc who put in more than 5 years in the same lab.

That said, the advisors (and other people) sure do get excited over new people. It generally seems to last about a year. :)

At 7:53 AM, Blogger Andrew said...

It depends where you are - in Europe, labs tend toward the 'old model' where techs/managers are lifers, grad students are long-termers (even though the PhD in Europe is still limted to 3-4 years), and postdocs are medium-termers. While in America, it seems to be as Ms.PhD has described. Also, Europe has a lot more job protections that make it impossible to fire someone, so you do get essentially permanent technician positions (though sometimes they like to move around within the department/university to get a change of pace).

At 7:48 PM, Anonymous Anonymous said...

It's interesting to hear how lab science works. I'm on the math/theoretical physics end, and here the max length of a postdoc has remained rigidly at 3 years. What has changed is the median number of postdocs has grown from 0 to 1 to 2. (Anecdotally, that is - I don't know the actual stats.)

The variety of experience is nice, but moving around well into your 30's gets difficult: hard if you have a partner, worse if they are also in academia, worse still if you're trying to start a family.

Incidentally, 1 year postdocs (which do exist) are miserable: you begin applying for jobs almost immediately after you arrive!

At 10:03 PM, Anonymous Anonymous said...

I started a postdoc in Molecular Oncology at the end of 2003. And the postdocs in my lab, as well as those in neighboring labs, are all. still. there. Nobody has moved on!!
Phds= a dime a dozen.


Stanford PhD

At 3:08 PM, Blogger Meredtih said...

I agree- but in my field it takes at least a 3 year postdoc to make enough progress on a project to tell a cohesive story and publish some quality papers. And even then, some people opt to do another postdoc. Many advisors in my field won't even take on postdocs unless they are willing to make a three year commitment, because looking for a job is almost a full-time endeavor during your final year. When I leave this summer for my "real job", I will have been in my current postdoc for almost 4 years, and that's pretty normal. I'm all for the idea of a shorter postdoc, but I'm afraid that not much would really get accomplished in my line of work on a brand new project in 2 years, while looking for a job.

At 2:53 PM, Anonymous Jane Chin, Ph.D. said...

I agree that postdocs should be shorter. Unfortunately, duration of postdocs is directly proportional to availability of job openings. If there are open spots at universities and industries for a postdoc, we may see an unprecedented productivity from postdocs motivated to nip that project in the bud and move onto better-paying jobs.


Post a Comment

Links to this post:

Create a Link

<< Home