Wednesday, November 25, 2009

Lessons learned in grad school, part whatever

This week I want to emphasize an important lesson I think everyone should learn before being awarded a PhD.

Do you want to get the right answer, or do you want to be obedient?

Lately one of my biggest concerns with ethics in science is the enormous pressure I see on grad students and postdocs, which leads them to fudge or even fake data.

I see this happening out of desire to be obedient, to be liked, and out of fear of being fired, or being wrong (or losing their visas and being deported).

So here is my two-step plan, because I think it's a major source of evil in science.

1. Do what you adviser says you should do, even if you're sure it's not going to work.

I say this for two reasons.

First, because YOU might be wrong about whether it will work or not.

Second, because then you can show them the data and show how obedient you were. And they'll never admit they were wrong without you showing them the data. They'll be much more likely to swallow their anger if they see that you did what they asked in good faith.

2. STOP doing what your adviser told you to do when it's clear that it's never going to work. Do what YOU think will get you the right answer.

There are two important lessons in this step.

First, knowing when to stop. This is a hard lesson and many people don't learn it until their postdoc is over and they're hunting for a new career. Don't be one of these people. Learn how to assess when you're making progress and testing possibilities, and stop and find another approach when you're just banging your head against a wall.

Second, knowing how to be brave and disobedient. This is a really hard lesson for most people in science, so there are options for how to go about it.

Doing what your adviser asks first is generally the safest route in this regard, although it might be the most inefficient.

Doing both your adviser's stupid idea and your awesome new thing at the same time can work for some people who are good at time management (not everyone can manage this).

Finally, doing your new thing at night or on the weekends when your adviser is traveling is the sneaky way. Notice that I did NOT say you have to ask them. DON'T ASK YOUR ADVISER. Just do it.

Most of the time, they will be overjoyed that you showed some independence and got the right answer. And if it's really a big deal, they'll claim is was their idea to do it your way all along.


What to do if you find out your adviser was wrong and they don't want to admit it even after being faced with data proving they were wrong?

That's a different blog post.

Happy pipetting, y'all. Oh and don't eat too much turkey.

Labels: , , , , , , ,


At 7:53 AM, Blogger daisy mae said...

i didn't know how to contact you via email, so you don't have to post this if you don't want.

but i came across this article:

and i thought of you.

and as a PhD student who is really, really struggling with her advisor, it made me cry.

i really really appreciate your blog.

At 8:51 AM, Blogger PhDamned said...

You are (in my opinion) 100% correct!

I think this is the best advice I got before grad school. My grad student mentor when I was doing undergrad research told me this exact same thing.

I think that hearing this from someone BEFORE I started my PhD was extremely fortunate for me and crucial to my success.

At 8:55 AM, Anonymous Anonymous said...

Very good post. I am now looking forward to your next post about the scenario when the advisor does not want to admit (s)he is wrong even after seeing the data.

At 12:00 PM, Anonymous Anonymous said...

Being an advisor myself, I can tell you that this is indeed the correct strategy, no matter how strange it is.

In fact, I expect advisees to follow my suggestions even though these suggestions are not the ones that will generate great results. They are the safe steps, ensuring that the advisee is doing some progress. It ensures some "safe" results, which will be rather boring but publishable.

The advisee should have the initiative to start thinking independently, create new solutions, and see how well they perform compared to the "expected", boring solutions.

Students should really understand that when they have all day to think about a problem, they can really do much better than their advisor, who gets to talk about 10 different ideas in a day for 10 different projects, and never gets the chance to think deeply.

At 9:45 AM, Blogger Ms.PhD said...

daisy mae,

That article is from 2000. I wish I had seen it. I don't remember though, maybe I did. Even as a grad student, I was convinced I would never do a postdoc. I was not that stupid. I knew it was a dead-end proposition.

Just don't drink the kool-aid. The difference between now and 2000 was that in 2000 there were other jobs. Now there are very few options that will let you "use" your science degree. Although, I disagree with the mentality that you're a scientist forever. I have a hard time imagining that when I've been out and away from the lab for 5 or 10 years that I'll still be a "scientist".

And I think most of my friends who left for other careers would argue that they DON'T use most of what they learned in their PhD training, and they certainly didn't need to spend so many years suffering just to get a piece of paper that says they have this degree that is almost irrelevant to what they actually do on a daily basis.

At 10:37 AM, Blogger biochem belle said...

I agree with all the way. This is the way things worked with my grad adviser. When I was more senior, I advised new grad students and postdocs the same.

The beauty(?) of my grad lab was that my adviser expected this (even if he didn't come out and say it) and paid little attention to what we were ordering or core services we were using (so long as it wasn't ridiculously expensive). This unspoken agreement fostered an intense independence as a grad student.

This sense of independence has been feeling repressed during my postdoc thus far. Part is because I took a postdoc in a very different field (with very different lab and department politics) using very different techniques. So I felt it was important to 'tow the line', so to speak.

Unfortunately that has not been terribly productive thus far. I had decided it was time to begin developing my own ideas, and this was further galvanized by advice from my graduate adviser. The issue I am encountering is that my postdoc adviser is a borderline micromanager. By this I mean that he does not oversee every experiment, but he is very involved in experimental planning and design and requests weekly progress reports from everyone. He also approves (or shoots down) every order. This makes it a little more difficult to do 'clandestine' experiments, unless we already have the reagents in the lab. Any thoughts?

At 10:48 AM, Blogger JaneB said...

As an advisor myself, I agree with the first anonymous. You are right - I make my suggestions in good faith and for a good reason. I am invested in you getting results, being productive, and will always start you on a 'safe' route. I hope you will develop independence - but also learn standard practices along the way.

In addition, there are reasons why staqndard protocols are standard. I had a student who always read protocols then modified them without trying the standard method first. He never grasped that a) because he had no 'standard' data he couldn't compare his findings with anyone elses and b) it was a total waste of his time (and my grant money, paying for chemicals and lab resources and salary) to try and reinvent with wheel without first showing that there was a real reason to spend time developing alternatives. So frustrating for us both!

At 11:15 AM, Blogger Ms.PhD said...

biochem belle- I SO feel your pain. I have had to do all of my research with zero money, either because my lab was broke or my advisor was a penny-wise, pound-foolish micromanager.

I sure can pick 'em!

My only advice is to beg, borrow and steal.

You can meet a lot of people this way and expand your network, which can be a happy side-effect.

And it's one of the only good things about having too many postdocs in the pipeline- you can usually find several other postdocs who are in the same boat in different labs, so you can usually barter your expertise or reagents to get what you need (and maybe some co-authorships on papers, too).

If things don't improve through hard work and communication, though, I would consider switching labs before it's too late. This lab does not sound like a great fit for you. It's really no fun doing science without support or resources.

Unfortunately, no matter what your resources, you can't get very far in this business unless your PI really supports you. And you're at a disadvantage for that as an XX. Sometimes you can win these people over- I'd say give it a year. If he doesn't warm up and start trusting you, GTFO and find someone who appreciates your self-directed efforts and gives you what you need to do your job. Don't get stuck in a crappy situation- it's much harder to switch labs the longer you stay.

At 2:10 AM, Anonymous Anonymous said...

I and a few other friends and colleagues (we are staff scientists) have often had to work on our "real ideas" secretly, spending the day doing the boss' stupid ideas which lead to incorrect results or correct but boring and insignificant results just because they are "safe" and the boss is only concerned with his appearances to his bosses. Then at night or on weekends we try to sneak in the more meaningful and riskier experiments we really want to do and exercise the intellectual independence that is the reason we got our PhDs in the first place.(rather than to become highly skilled professional lab drones to further the career of power-hungry managers).

I've had varied success with this approach. While it is rewarding to be able to (secretly) do the work YOU want to do and see it come to fruition, it can only go so far without being "official" especially if it requires expensive resources that you can't bootleg or borrow under the table. And even if your secret project ultimately succeeds scientifically, what do you do about publications? the boss may be happy to have gotten this extra work out of you. Or, they may be furious that you were doing this without their approval. Or, they may be furious that you were doing this even if they are excited by the results but simply because if you're gonna work extra (without overtime pay since you're salaried) you should be doing it for "their" stupid ideas and not doing your own thing.

Honestly, I had much more freedom in grad school than I do now as a post-postdoc staff scientist even though I'm so much more technically- and scientifically mature now from having 10+ more years of experience above the PhD (spent doing hands on science not sitting behind a desk)

At 11:27 AM, Blogger Ms.PhD said...

Anon- Absolutely. This only works up to the point when you need more resources or publications. Most PIs will be happy to take credit for your stuff if it's working, even if they told you not to do it. But not all. Some are still stubborn, selfish, threatened, or incapable of objectively assessing the data you're showing them.

That's the thing about freedom... science isn't free. You're only free so far as you can afford it.

At 8:23 PM, Blogger Boyan said...

This discussion has a lot of victim mentality associated with it that is rather disturbing. If you want money you should not be in science. You should not be asking "are lawyers worth 10x what I am worth", salary should not be the primary motivator of what you do. You do it because you are intensely curious, because you want to know, because you are persistent, because your heart tells you there is no other way.

Yes, I did a postdoc and never felt it was a bad thing. If your prof is doing his/her job, they will give you a taste what academia is like that is usually not open to grad students. It is this bit that showed me that academic life is absolutely not for me, that I cannot deal with the shameless self-promotion it takes, with the huge spin that minor results require in order to be noticed and attract grants.

Financially, the post-doc has cost me dearly. Three years at $30k vs. $70k or more for starters. But even more importantly, it cost me three years of appreciation of stock options in the late 90s, that probably translate to several $100k of "missed" income. Am I sorry? Absolutely not, I would do it all over again in a heartbeat because a) I have never been motivated by money b) there is a lot of value in finding out who you are and what makes you tick.

My best advice is to love what you do intensely, do it only because you love it, and success will come. At the top decile ANY profession that requires a Ph.D. will pay you more than enough to lead a comfortable life. And if you love what you do, you will rise to that decile. On the other hand, if money is your major measure of self-worth you might as well quit early and go into law or finance, where $$$ is the universal measure of everything.

At 11:05 PM, Blogger cookingwithsolvents said...

The system is strange and varies tremendously from PI to PI.

One can make a significant argument that the day you stop listening to your boss and do the experiments YOU know are best for your project (and can defend the direction and methods of scientific inquiry) as the day you begin to earn your PhD. You and your boss will be working together to answer questions (just 'cause you are working towards/have a PhD doesn't mean you know everything about everything, of course).

Dealing with a PI that doesn't see science that way, well, that's politics and good luck.

At 9:04 AM, Anonymous Devin said...

That is some solid advice. There are some similar discussions, and advice, over at GradShare. Anyone interested in this should check it out.


Post a Comment

Links to this post:

Create a Link

<< Home