Layla Parast, University of Texas at Austin, has some advice for beginner researchers. The small things, done well and reliably, really do matter. She writes:

Every year, I talk to undergraduates who are interested in “doing research.” This is, in general, a very good instinct. Research is one of the best ways to actually learn what statistics is, beyond problem sets and exams. It is also one of the fastest ways to figure out what you do, and do not, enjoy.

That said, there is a persistent mismatch between what students think research is and what it actually is.

A common misconception is that research is just harder homework. It is not. In homework, someone knows the answer. In research, no one does. That is the entire point. Sometimes the problem is too hard, sometimes the idea fails, and sometimes the result is that there is no result. This is not a bug; it is the job.

So, if you are an undergraduate trying to get involved in research, let me offer some advice. None of this guidance is complicated, but it is apparently necessary.

First, engage. Go to talks. You will not understand most of what is happening. That is fine. In fact, that is expected. Look, I don’t even understand those talks. The goal is not immediate comprehension; it is exposure. But if you go, actually go. Listen. Try to follow something, literally, anything. If you ask a question, ask it because you are curious, not because you want to sound clever. (These are not the same thing, and faculty can tell the difference immediately.)

Second, figure out what you like. Not in some abstract, touchy-feely sense, but in a very practical one. What kinds of tasks do you enjoy? Writing code? Proving things? Cleaning messy data? Reading papers? Research is made up of many small, sometimes tedious components. If you hate all of them, this may not be your path. If you like even one of them, that’s a starting point.

Third, and this is the one I wish I could put in bold, underline, and repeat three times: be reliable. If you say you will do something by Friday, do it by Friday. Not Friday evening with an apology, not Saturday morning with an explanation. Friday. From a faculty perspective, I can teach you almost anything technical. I cannot teach you to meet a deadline. Reliability is the entire currency of early research opportunities.

Closely related: do small things well. Early on, you are not being asked to solve open problems. You are being asked to run code, check a derivation, reproduce a figure. These tasks are not trivial. They are tests of trust. If you do them carefully and completely, you will be given more interesting things. If you do them carelessly, you will not.

Another small but important point: communicate like a professional. This does not mean writing formal emails with elaborate greetings. It means being organized and respectful of time. If you run into ten small issues, do not send ten emails. Keep a list. Think about them. Try to solve them. Then send one clear, structured message. Even better, bring them to a meeting. And before asking “why did this fail?”, spend a little time trying to answer that question yourself. When you do ask, say what you tried. This turns a vague question into a productive conversation.

Also: write things down. Like, with a pen (or pencil). On this thing that we call – paper. Let’s say it together, PAPER. You will forget what you did, why you did it, and what worked. Everyone does. The difference is whether you planned for that.

Now, a brief comment on ego. Leave it at the door. You will be wrong frequently (also, I will be wrong, frequently). That is not a reflection of your ability; it is the nature of the work. What does matter is honesty. If you did not finish something, say so. If you do not understand something, say so. Pretending otherwise does not make the problem go away—it just makes it harder to fix later, often for someone else, like me.

Finally, understand what the process looks like. Early research is mostly learning and failing. Often in that order, sometimes simultaneously. You try things that don’t work. This is, counterintuitively, a good sign. It usually means the question is interesting. Over time, if you stick with it, you move toward more independence: identifying ideas, proposing approaches, owning pieces of the project. But that transition only happens if the foundation is solid.

All of this boils down to something very simple. Faculty are making an investment when they take on a student, undergraduate or graduate. Students are making an investment in return. When it works well, both sides benefit.

So, if you are looking for research opportunities, you don’t need to be the smartest person in the room. You do not need to know everything in advance. You do need to show up, do what you said you would do, think a little before asking for help, and take the work seriously.

It is a surprisingly effective strategy. And, for reasons that remain unclear, a surprisingly uncommon one.