Read on Twitter
I was asked to share more on this—finding in grad school that research is harder than coursework—and things that helped me learn to be a better researcher.
I was a theorist—I can’t tell how much will apply if your work revolves around experiments/observations.
Here goes!
https://twitter.com/astrobassball/status/1354062039887601665
I was a theorist—I can’t tell how much will apply if your work revolves around experiments/observations.
Here goes!
https://twitter.com/astrobassball/status/1354062039887601665
These were the big shifts I faced when trying to do research vs. coursework. I’ll go through them separately.
1. Structuring time
2. Breaking down the problem
3. Asking for feedback and help
4. Validation and self-worth
5. Embracing uncertainty + ignorance
6. Take ownership
1. Structuring time
2. Breaking down the problem
3. Asking for feedback and help
4. Validation and self-worth
5. Embracing uncertainty + ignorance
6. Take ownership
But first, an aside: much of this will require shifts in thinking and daily routines. It will require the creation of habits conducive to research progress, and curbing some that aren’t.
Something I took away from Aristotle’s Nicomachean Ethics in high school:
“We are what we repeatedly do. Excellence, then, is not an act, but a habit.”
And “Every habit begins with a single action.”
Start by doing the thing once. Then do it again the next day. Rinse, repeat.
“We are what we repeatedly do. Excellence, then, is not an act, but a habit.”
And “Every habit begins with a single action.”
Start by doing the thing once. Then do it again the next day. Rinse, repeat.
Another piece of context. I was insufferably arrogant when I began grad school. I aced courses w/o attending class, and foolishly thought that meant something. Terrible attitude. My ego blocked me from being an effective researcher, and made me a bad peer & a poor collaborator.
It took lots of self-inspection & time to change. Had a lot of growing up to do. For most of us, grad school is early adulthood, formative for the rest of adulthood. Please give yourself the time & space for your own growth, to navigate & learn from relationships, etc.
Okay, back on topic.
1. Structuring time.
You may have large blocks of time between classes and other obligations. Whereas this was “free time” in college, it’s time at work as a grad student researcher.
1. Structuring time.
You may have large blocks of time between classes and other obligations. Whereas this was “free time” in college, it’s time at work as a grad student researcher.
You definitely don’t need to work 60 hours/week to be a productive researcher. But if you’re only doing research 10 hours/week as an early-career researcher, that could be a tough formula.
You may need to figure out when your most focus-friendly hours are. Is it after your morning coffee? Is it late at night? Try different things, but commit the time.
Do you have a distraction-free work space? Do you need to block out certain websites or apps?
Sometimes, I delete the Twitter app from my phone for a few days, when I notice myself using it too much. There are browser plug-ins to help limit time on websites.
Sometimes, I delete the Twitter app from my phone for a few days, when I notice myself using it too much. There are browser plug-ins to help limit time on websites.
2. Breaking down the project into specific components.
Research comes down to papers, and papers come down to punchlines. (Many well-cited researchers have refined the art of turning results into catchy paper titles.)
Research comes down to papers, and papers come down to punchlines. (Many well-cited researchers have refined the art of turning results into catchy paper titles.)
But in the beginning, I usually think about The Plot. “What is the plot that will best illustrate the answer to this question?”
Then, “What do I need to make this plot?”
What methods, equations, principles do I need to learn and apply? (Literature search.)
Then, “What do I need to make this plot?”
What methods, equations, principles do I need to learn and apply? (Literature search.)
What calculation, code or techniques do I need to implement and perform? (Tool/code development)
What tests will give me confidence that things are working?
What tests will give me confidence that things are working?
Break down each of these into steps.
Set reasonable, but stretchy, goals for each one, and share with your group or supervisor often.
Set reasonable, but stretchy, goals for each one, and share with your group or supervisor often.
3. Getting feedback and help.
In coursework, you often work through everything—you may even write up a clean copy by hand or with LaTex—and then submit.
In collaborative research, I like something a former coworker said: “Ask for feedback when you’re 20% done, not 80% done.”
In coursework, you often work through everything—you may even write up a clean copy by hand or with LaTex—and then submit.
In collaborative research, I like something a former coworker said: “Ask for feedback when you’re 20% done, not 80% done.”
Before going too deep down a rabbit hole or a specific approach, trust the experience of your advisor and group. They may know shortcuts, or be able to suggest alternative approaches.
Take a deep breath. Repeat after me: You don’t need to figure it out alone.
I’ve been the “single author” guy churning out results in solitude. I’ve also been on, and led, collaborative projects. You will get farther, and more people will see the value of your work, with the latter approach.
And this will mean asking for help. Again, if you were the sort to ace coursework without going to office hours, this may be uncomfortable and humbling.
4. Validation and self-worth.
Coursework comes with a set timeline and grades. Research will often feel like a marathon.
You share results with your collaborators, and you go back to your desk and keep going until you have enough to write a paper.
Coursework comes with a set timeline and grades. Research will often feel like a marathon.
You share results with your collaborators, and you go back to your desk and keep going until you have enough to write a paper.
You may get a lot of encouragement in this process. You may not. It’s important to have a source of validation and motivation.
Your collaborators may not be the sort to leave you feeling like you had a great, productive week. You may have weeks that feel like total busts.
Your collaborators may not be the sort to leave you feeling like you had a great, productive week. You may have weeks that feel like total busts.
You’ll have to learn to manage that—what defines a good day / week / semester / year for you?
How will you measure progress, and what defines success?
Your answer will be different from mine, and your peers'. I strongly recommend spending the time to reflect on this regularly.
How will you measure progress, and what defines success?
Your answer will be different from mine, and your peers'. I strongly recommend spending the time to reflect on this regularly.
5. Embracing uncertainty and ignorance
Most homework problems have been vetted by the teacher to have a tangible solution. Research is by its nature uncertain, and often has non-results that are ambiguous or plain dead ends.
Most homework problems have been vetted by the teacher to have a tangible solution. Research is by its nature uncertain, and often has non-results that are ambiguous or plain dead ends.
(You do develop with experience a sense of what is answerable and how—hopefully your research supervisor has thought this part through before giving you this project.)
Which is to say that it’s not only okay, but expected and even exciting, to go to your collaborators saying “I found something totally weird.” or “This looks so wrong and I can’t figure it out.”
For many folks I know, this is the BEST part of doing research.
For many folks I know, this is the BEST part of doing research.
One of my advisors is pretty stoic, and hard to read.
It initially scared the crap out of me when he would say during a research meeting, “I have a question….”
Until I learned that he wasn’t asking me to test me, but that he genuinely didn’t know and wanted to.
It initially scared the crap out of me when he would say during a research meeting, “I have a question….”
Until I learned that he wasn’t asking me to test me, but that he genuinely didn’t know and wanted to.
No one knows the answer!!
You might be the first to know when you figure it out! And then you'll be the one teaching your advisor.
You might be the first to know when you figure it out! And then you'll be the one teaching your advisor.
6. Take ownership
The piece of advice that gave me the confidence I needed to lead my own research was: “You’re a good student. Take ownership of your projects.”
The piece of advice that gave me the confidence I needed to lead my own research was: “You’re a good student. Take ownership of your projects.”
It helped me accept that I don’t have to worry about taking the wrong approach, that in all probability no one else is going to try approach X unless I do.
It was a freeing realization, and shifted my approach from “trying to get the answer” to “trying different things.”
It was a freeing realization, and shifted my approach from “trying to get the answer” to “trying different things.”
PHEW!
Okay, that’s about it. Let me make a couple more remarks and wrap up.
Okay, that’s about it. Let me make a couple more remarks and wrap up.
A former boss (who doesn’t have an academic background but has an affinity for hiring PhDs) likes to talk about the ability to deal with “complexity and ambiguity.”
Obviously, not all PhDs have that ability, and many without doctorates do. But I think it’s a skill set that’s developed through the experience of doing independent research.
You pose a problem to them, and they have the intuition and the technical competence to formulate and execute (with statistical sensibility & decent code) different approaches and come back with their findings—be they answers or more questions.
If you emerge from your grad school experience with that ability, I would say that’s a resounding success and accomplishment.
And that’s so much more important than how many papers you wrote or how many fellowship offers you got.
And that’s so much more important than how many papers you wrote or how many fellowship offers you got.
Anyway, those are the difficulties I faced in learning how to do research, and some things that I thought might help if you face similar hurdles.
I’m not an expert, and don’t claim to be.
If this thread adds to your perspective, that would make me happy.
I’m not an expert, and don’t claim to be.
If this thread adds to your perspective, that would make me happy.
