I am postdoctoral fellow working on computer science and bioinformatics, and it seems I might have the possibility soon to supervise some undergraduate internship students. Internships should last around 3 months.

To do this, first I have to design some undergraduate internship project proposals to make available for these candidate internship students.

I never did this previously, so I do not know very well how to design these project drafts. There are many question marks: what should I expect from these students? What should I expect from these projects? what amount of work should I expect? What skills should I suppose the students have before joining the project? What should be the goals of these projects, both for the students and for the advance of science?

In the end, in case you were a potential supervisor, what's a good way to design an undergraduate internship project proposal?

1 Answer 1


The number one thing that I would advise you to expect is the unexpected. The skills and performance of undergraduates is almost impossible to predict, because their work with you is likely to be one of their first deep technical experiences. Among the key implications of this:

  • their true skills and background vary greatly and are often not well indicated by transcripts (and they often have little ability to self-evaluate)
  • their motivation and ability to act professionally varies wildly and is very difficult to judge before they are actually working on a project (they often do not really know themselves)
  • there is often little possibility of obtaining a meaningful reference for them

Given this, I have found that the following are very useful to do when designing a project for an undergraduate:

  • Do not put them on anything critical path. Choose an interesting side project that you've wanted to get around to, but won't be sad if nothing happens on.
  • Create a set of milestones that starts with simple "warm-up" work, and then builds sequentially. Make sure they will see a concrete result that they can feel proud of at each milestone.
  • A good rule of thumb is to plan assuming that 1 week of work for you will be 1 month of work for the undergraduate. A really good one, though, may progress much, much faster, though, so...
  • Also add a set of "stretch goals" that will give a really good undergraduate a chance to do some really cool things, rather than running out of work after the first few weeks.

Every undergraduate is a roll of the dice. Many will simply waste your time. Many will be marginal in value at best. But the good ones really make up for all the rest, and may become your future colleagues.

  • 1
    I know this sounds kinda cruddy, but maybe 50% of these times I see too many grad students focus too heavily on the very temporary undergrads "helping them" in a lab, to their own despair. The situation usually severely cuts the productivity of the grad student, or in this case the post doc. I'm in no way insulting undergrads in labs, I've met many who have made immeasurable contributions. But generally they are not around long enough to get enough training to make progress. The situation usually stems from the PI w little to know management experience. Just a thought
    – rhill45
    Commented May 16, 2015 at 3:31

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .