choosing a research topic

[eco_env]

My advisor is giving me complete freedom to choose my research questions, system, topic, etc, but it seems to me that it would be most productive to do work that branches off of/relates to previous work in this lab. I was thinking I'd read some of this lab's work/proposals and see what interesting questions I get from that. is that the wrong approach? how did other grad students choose a research question that worked out?

[robot_hamster]

I was always taught that it is best to pick an area of interest first and then look to see what has been done in that area. As you read through papers, you may say to yourself "okay, they did this and this, but why didn't they do this?". You keep going from there until you see somewhere that you feel is lacking and has a question that is worth trying to answer. At my school, you have to write up a proposal anyway that includes a literature review, so you're given a chance to explain why this research is important and demonstrate how other studies failed to address the question you wish to answer. It sounds a lot easier than it really is! But I think that you have a good idea, to look at what others have done in your lab and see what has been left unanswered.

[Eigen]

I agree with the previous post- pick a general area of interest, and then start reading papers. Lots of papers. I spent almost 2 solid months when I started reading papers to get a handle on the area, and what had been done- it's slow going, but you see where there's potential for new projects, as well as getting a better since for what really interests you.

I would start with reviews- see generally what's been done, and how it fits together, and then branch out into the related literature.

It's definitely not a bad idea to see what has been proposed/done in the lab in the past, it will give you a good framework to start from.