Advice on Work Habits
This is a running list of advice accrued through the years. As with all advice, take what you need and try out a new idea if it sounds useful.
General productivity tips
Optimize your time management skills: This is arguably one of the most important skills you will hone throughout your career. As you gain experience, you will become more familiar with how long things take. In my experience, things often take about 1.5-2x longer than predicted. Prioritize, plan, execute, and check. Some strategies
make a list of priorities and "eat the frog" first (i.e., do the least appetizing thing on your list first)
keep track of how you spend your time to see where it goes and adjust accordingly (e.g., mark your calendar with how you spent your day; tally up your hours spent periodically, say every semester or so)
learning when to say something is good enough and when to say something still needs work (e.g., will this work benefit my short/medium/long term goals?)
Aim to work smart: Make the most of your time while doing research. That means minimizing distractions (e.g., from social media, emails, meetings). It could also mean maximizing the chunk of time in the day to work on a task that needs intense focus. It could also mean spending time to make a nice script to automate a task you think you will have to do multiple times.
Figure out your zone and workflow: Everyone has different work habits and it’s important to figure out what works for you to get to an optimal level of productivity. That could mean making to-do lists, having a calendar, using a task manager. Find what works for you to keep track and keep record of all the details.
Maintain regular work habits: Research is hard work. A PhD is hard work. You will need to be prepared to dedicate many hours. Empirically, excellence comes from sustained and persistent effort. Good science happens with consistency and hard work. You can be successful with a 9-to-5 schedule but this requires excellent focus and time management (most people, PI included, do not have such excellent time management skills). Of course, the more intense work hours you put in, the better the end product will generally be. From experience, typically this is on the order of ~55 hours per week.
Be flexible: All sorts of things happen (e.g., a deadline, a poor reviewer report, a huge result), so it is important to be flexible and keep a steady pace. Sometimes you may need to dedicate your weekend(s) in order to meet a deadline. We will do our best to avoid this, but keep this in mind.
Maintain a regular working schedule: I highly encourage maintaining a regular work schedule. It is useful to be in phase with everyone else like faculty, staff, and other students. Otherwise you may miss out on useful conversations and opportunities.
Continually think about improvement and feedback: This includes technical knowledge, scientific communication, and writing! No one becomes a good scientist overnight. Rather it is the accumulation of many small improvements that compound over time. Examples include making a connection at a conference, learning a new way to analyze data, learning more tricks, solving problems faster. Perhaps you had one way of doing things a while back, but technology has changed and new tools have been developed so it might be time to revisit and revamp what you had before.
Identify and connect with mentors: There are many types of mentors, and each type of mentor can be helpful in different ways. It is usually helpful to have several mentors along different stages of their careers so you get to learn the bigger trajectory in how to navigate your time here
Become the expert of your project: eventually you will know more than your advisor about the intricate details of your research.
Habits of a Good Scientist
Record as much as you can: Research is incremental. You will try lots of little different things, run many variations of the same kind of calculation in order to understand the problem and understand your results. Keep track as much as you can! Whether in a notebook, excel spreadsheet, markdown, whatever. Make sure you (or someone else) can look back many years from now and be able to reproduce what you have done. This includes seminar notes, intermediate data, analysis scripts, etc. You should always ask yourself, should I record this, and if your question is should I record this (and organize it), the answer is almost always yes.
Be "hungry" and ambitious: Take advantage of any and all resources you have at your disposal.
Be ambitious in ideas/execution, conservative in interpretation: ideas are the first step in any research, carrying out the research is usually the bulk of the work; you will find things are often more complex than anticipated. Be sure to also keep track of big ideas you brainstorm; let them marinate and refine. The significance of a result needs to be carefully considered and phrased, accompanied by a thorough understanding of the literature.
Be courageous: many times we will not be able to anticipate exactly what will happen when we embark on a project. Sometimes ideas will flop or fail. But that is also the exciting part of research and a little risk tends to be a part of it.
Be honest: progress in science depends on people reporting results truthfully and in good conscience. No good comes from sweeping things under the rug or ignoring strange parts of the data; it can also bite you back later.
Be your own harshest skeptic: This is of course something to develop over time, but the goal is to be able to first convince yourself of your results so that you can convince others. Be critical of your results. Does it match your physical intuition? Is it the right order of magnitude? etc.
Know your fundamentals: Cool technologies and scientific advances are only enabled by a strong understanding of the fundamentals. Be on the constant lookout for how to improve your fundamentals. Read seminal papers and textbooks. Revisit what you have learned before, discuss it with people.
"The devil is in the details": Pay attention to details, they matter.
Quality over quantity: Of course the more papers and presentations the better. But our priority is quality. We aim to push out publications of high quality and high impact. It is simply not worth it for ourselves or the scientific community to compromise quality for the sake of quantity.
Know when to ask for help: No one person knows everything. It is ok (and encouraged!) to reach out for help, especially if you are stuck. It is helpful to have a well-formed question or problem first (i.e., not “it doesn’t work” or “it doesn’t make sense” but something more specific like “I tried this and this. I found this in the literature, which is a different result and it is possible the discrepancy could be for this and this reasons but I am not sure”).
Have an online presence: An online presence is particularly useful in the 21st century- for establishing your reputation, for making new connections, for applying to jobs. Be sure to have some kind of webpage that you curate. This could be LinkedIn, Google Scholar, ORCID, ResearchGate, a personal website (e.g., Weebly, Wordspace are free), Github, etc. or a combination of all of them.
Develop a skill set: While there are many philosophies on what exactly this entails, I find it is useful to have two frameworks in mind: 1) aim for a “T-shape” where you have broad and shallow knowledge across many areas and deep in a select few; 2) aim for 2-3 orthogonal skills so that you have the flexibility to approach problems and come up with creative solutions (example: expertise in solid-state physics, coding skills, science communication)
Keep up to date with the literature: a literature search is critical for any major project. You must know all the relevant historical papers and the recent ones as well. A thorough literature search is the foundation for any successful project. For graduate students, you are expected to know how to do a literature review by the end of your first year
As you become more familiar with key journals and players in the field, you can set up RSS alerts, Google scholar alerts, etc.
Last updated