Before diving into "real research", there are some students who prefer to learn bare minimum prerequisite courses/material and then start off reading survey papers and develop their skills.

There are others who prefer to go through courses from Introduction to Advanced to Independent Study before finally diving in.

The former would probably face problems of fundamentals while the latter spends too much time learning things which might not be completely useful.

At what point should one (Assume, if necessary, that I am talking about STEM) decide to dive into research while making sure that one is neither being too hasty nor slow?

Take for e.g. that a person wishes to code a software in Python which does engineering calculations. He would either read something like Intro to Python and then directly start coding or he could also read documentations of other Math libraries, similar libraries in C/Fortran, study coding efficiency and thumb rules and then start.

How do you prevent yourself from taking up too little or too much time to begin?

  • 2
    Why the vote to close?
    – JeffE
    Apr 1, 2012 at 7:09

3 Answers 3


As aeismail says, there's no single answer. As with most things in life, it's hard to know when you're "ready" to do something. The best way is to jump in when you feel like you don't need to prep any more, and then be prepared to shore up weaknesses as you spot them.

Over time, you'll learn how to prepare yourself to embark on a new research topic: some people like to start solving a problem first, and then return to the literature when they get stuck. Some people like to read a few key papers to get a sense of the main open problems and techniques, and then go off and play.

The only important thing is that you do something, and not just wait on the sidelines for a mythical feeling of "readiness" to emerge. Be fearless, like a 2 year old who doesn't know enough to be afraid :)

  • 1
    +1 The only important thing is that you do something, and not just wait on the sidelines for a mythical feeling of "readiness" to emerge
    – user107
    Apr 2, 2012 at 16:30

I've never understood the dichotomy between “preparing to do research” and actually doing research. I have always learned best with a target problem in mind. I always have to learn new fundamentals to solve any new problem. Most of the time I spend on research is "wasted", and most of what I learn is "useless", and that's okay.

The difference between just solving interesting problems and doing "real research" is pretty small. In both cases, you're completely ignorant at the beginning; what distinguishes "real research" is that everyone else in the world is ignorant, too. As long as you're comfortable working with no hope of finding your answers in the back of a book, you're as ready as you'll ever be. Jump in!


This isn't a one-size-fits-all problem. People should move into research when they're ready to do so, and in consultation with their advisors, when appropriate.

That said, the approach I'd tend to advocate is to ramp down classwork while ramping up research. In that sense, the student controls the pace at which she learns, and can adjust the selection of coursework as time goes on to support or to complement the research work. Moreover, there's generally the assumption on the part of the advisor that the first few months aren't going to feature a lot of useful scientific results; they'll mainly be spent learning techniques and tools and basic concepts and understanding.

So the way to figure out if one is ready to start research is by doing some "low-hanging fruit" problems: if the student can handle the basics, then she can start moving on to the rest. If not, then at least she has a better handle on what she needs.

  • How do you gauge that you are ready? I mean, its easy to feel ready when you are not and the other way around. I tend to keep reading trying to convince myself that I'm still not ready. Consultation is one way. Is there any other?
    – user107
    Mar 31, 2012 at 21:05
  • @Nunoxic: Yes: doing some simple research-style problems. (See my revised answer.)
    – aeismail
    Apr 1, 2012 at 6:39

You must log in to answer this question.