How to Win at Science

February 6th, 2009

Warning: this post is practical advice for the individual and risks breeding cynicism toward science.

I am a tenured professor of astronomy at the University of Wyoming, a category I research university where I enjoy a relatively low teaching load and decent pay, and where I lead a research group that has brought in close to two million dollars in grant money in the last seven years.   I was a double major in electrical engineering and space physics at Rice University (turning down an acceptance at Caltech), did a PhD in astronomy at the University of Texas at Austin, and did post-docs at Lawrence Livermore National Laboratory and Kitt Peak National Observatory.   I have been on both sides of the job thing, for postdocs and faculty jobs, and also both sides of graduate school admissions.

I worked hard to get where I am and followed my love for astronomy even when other paths would have been easier and more profitable.   I am something of an idealist, and have been my entire life.   This doesn’t mean I am stupid or naive or can’t see how things work.

Read the following keeping in mind I believe it is intrinsically valuable to learn how our universe works, that science has show again and again that it is the best and most reliable tool for doing so, and that I think people are happier and more effective doing what they love to do on a daily basis.   Also I think it is important to have at least five qualities to pull off a winning career, but also a plan like the one I am going to describe below.

Disclaimers and background out of the way, let’s get practical.

Let’s say you’re a smart person who loves science and has those five qualities and wants to “win at science.”   That is, have a career successful enough to allow you to land the increasingly elusive permanent job and actually do science every day?   With the problems of the academic Ponzi scheme, getting that permanent job can be really hard.

But there are some relatively easy tricks to further success and to win.

Most people don’t think about this seriously.   It’s only about the next hurdle, e.g., landing the bachelor’s degree, getting the PhD, getting the postdoc, etc., and without serious planning these steps may be difficult.   It is akin to the problem writers have.   At first it is about finishing something, then about selling something, and after a while you realize you’re having a career, but it may or may not be a good one.   We’ll start with some general schemes and move on to some things specific to astronomy.

Generally, find out what things count most in the job market and do them.   Do them even if it means skipping other opportunities.   Get invited to give a big review talk at an international meeting?   By all means do it if it doesn’t take too much time, but skip it if the time will cost you finishing a paper.   One refereed publication on your curriculum vita (CV) generally counts more than the recognition of being an international expert.   The people with the permanent jobs are not usually people in your field and won’t care.   They will be counting papers on your CV.

And since the number of papers on your CV is one of the first things people look at, maximize them.   Don’t write a single long paper when you can get away with writing two shorter ones.   There is a time and place for writing the large comprehensive paper — one time is when it may be required is for your PhD project and to get some notice in the field (good for reference letters and getting hired as a postdoc), and the other time is after you get tenure.

Also you need to specialize in something you’re good at to maximize production.   This is a flaw of my own career.   I do everything in observational astronomy from radio waves to X-rays, from imaging to spectroscopy, from spectropolarimetry to spectral principal component analysis.   There are some advantages to this approach, but big drawbacks.   I will talk about this again in the context of funding.   Anyway, figure out what you like to do and what is relatively easy for you.   One thing ideally, but because of the dangers of being too limited it is better to pick 2-3 things that are complementary.   Maybe some kind of experimental technique, some kind of analysis technique, and some kind of modeling, that can be used together.   Become THE RECOGNIZED EXPERT.   This will get you invited to join grant proposals and get away with doing something easy and quick for you that will get you added on other people’s papers.

If you have a range of interests, find out which of those interests has the largest piece of the pie.   Subfields with more people have higher citation rates and usually demand a larger share of grant funding.   It’s hard to get funded or cited if you work in an obscure topic.

Pick your advisors/bosses and collaborators carefully.   Pick a productive advisor.   Research their publications and pick someone publishing a lot, not someone who is dead wood.   You will learn how to work efficiently and be pressured to be productive.   Some collaborators are more of a drag on a project than an asset.   Some always give good and timely feedback.   Some are black holes.   Figure out who is who and make adjustments as necessary.

While I think it is critical to develop original skills and learn how to do things for yourself, a lot of the time it is faster and easier to let someone else to the tedious work.   In astronomy we now enjoy large databases of images and spectra that are public, through projects like the Sloan Digital Sky Survey (SDSS) and the Hubble Space Telescope Multi-Mission Archives (MAST).   Sometimes you need to select a sample of objects, design an observational program, propose, wait, get the data, reduce the data, and do the analysis.   Sometimes you can just download the data and do the analysis, saving years.   Sometimes what you want isn’t available, but it is increasingly available these days.   I also remember being amazed at what can be done with published data — most scientists focus on only one small aspect of a dataset, and there may be more important results still hidden there, just waiting for someone else to come along.   Or by combining multiple published studies, a meta analysis, clearer results may be readily available.   Meta analyses are often very citable.

For permanent positions, getting grant money is often important.   Make sure you get experience writing grant proposals as soon as possible.   Help your advisor or boss as a grad student or postdoc.   Find out if you’re allowed to write your own grants as a postdoc.   Say yes to opportunities to review grants, and if they don’t come your way, find out who the grant officers are and email them to tell them you’re interested in being a reviewer in the future.   This is an absolutely golden way of learning how to write good grants.   It is the equivalent of being a slush reader for a science fiction magazine.   It becomes really easy to see what works and what doesn’t and easier to emulate the former.

Find out who in your subfield is regarded as the best and study their work.   Emulate their techniques and approach when possible (don’t copy their studies as they’ve already been done, of course).   Do complementary projects when possible, and confirm or invalidate their ideas (if you find a problem, don’t be an asshole!).   Do this well, and you’ll end up a collaborator at some point, or at least get cited by them a lot.

Don’t take negative referee reports or other criticisms personally.   Getting mad at people or having personal vendettas is rarely productive.   It is easy to blow an entire afternoon complaining about someone or a referee’s report or how a review panel didn’t understand your great ideas.   Swallow all that and make the revisions.

A few things now specific to astronomy. It’s faster and easier to start doing observational work as a student than theoretical work, at least if you’re sufficiently careful enough not to make too many mistakes.   If theory is your love, do it, but for me it was not what I loved.   Note that some of the most productive and respected people in astronomy are theorists, and if you’re a genius or you can can develop a theoretical niche it is possible to be fast.   Chandrasekhar didn’t need my advice.   Also a theorist doesn’t have to write a telescope proposal or wait for clear weather to make progress on their project, which is a plus, but they also tend to have a harder time finding grant funding.

An easy way to be a good, innovative observational astronomer is to be able to work at multiple wavelengths.   That is, there are usually obvious studies to do following up one project in say, the optical, with new observations in the X-rays.   Specialize in wavelengths where there is easy grant money.   Optical and radio astronomers have it tough since most projects with ground-based telescopes don’t offer funding.   Space-based projects often come with analysis money, so most X-ray projects, far-infrared projects, ultraviolet astronomy, etc., if accepted, are prefunded.   Other scientists often get projects accepted and then have money to hire a postdoc with expertise to do the technical work, so these are great areas to work in from a practical standpoint.

Finally, one last suggestion.   Instrument builders in astronomy are always in high demand and developing this skill is a great way to get a permanent job.   It is risky, however, as instruments can be slow to build (taking more years to finish a PhD), and don’t usually result in a lot of papers.

Summary

1. Maximize your publication rate.   Write a larger number of shorter papers.

2. Specialize.   Find a few things you can do better and faster than other people.   This will result in more papers, and more opportunities to collaborate and score easy co-authorships.

3. Pick a well funded subfield large enough to ensure a higher than average citation rate and grant opportunities.

4. Pick advisors/bosses/collaborators who are productive, and dump the slow, lazy, and unhelpful.

5. Work with public data when possible.   Meta analyses, well done, are highly cited.   Do your own experiment or observation only when necessary (it’s slow).

6. Develop grant writing skills.   Help others when given the chance.   Say yes to, or volunteer for, grant review panels.

7. Identify the best in your subfield and study their papers.   Adopt their best techniques.   Do complementary projects.   Collaborate with them if possible.

8. Don’t take things personally.   Bull sessions to complain about bad referees, ignorant review panels, or that idiot who wrote a paper critical of your work may be satisfying in the short run, but they can steal time.   Channel that energy into revising the paper or proposal so that even a total idiot will see its brilliance.

And specific to astronomy:

9.   It’s generally quicker and easier to develop skills in observational astronomy, and there’s also more funding opportunities.   Unless you have love or special talent for theory, skip it.

10. Specialize in 2-3 observational techniques/wavebands and don’t try to do it all.   Use radio and optical archival data as much as possible and focus on writing proposals that bring both telescope time and funding simultaneously (e.g. space-based telescopes like Hubble or Chandra).

11. If you’re interested in building instruments and can do it well, you will be in demand.   The path is riskier, however.

And now it is time to stop blogging and get back to revising a paper.

Share/Bookmark

Tags: , , ,

You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.