Assume that one has a skill (assume, programming) and is one of the very few in the department to have it. The other students need a bit of this skill for their projects and you help out when you can.
Where do you set the limit for how much effort/time you spend on helping others this way?
- You get insight into other students' work.
- It's a welcome break from our own research
- Maybe earns you a second/third author for setting up the experiment
- Networking !
- Effort/Time spent
- You don't necessarily improve your skills (For instance, coding in Python for someone else doesn't augment my Python skills by much. What I do might be really routine)
- You tend to have a soft commitment towards that project. For instance, if I start working on it as a favour, it doesn't really come off well if I leave 'em midway.
EDIT: I do enjoy the work so long as it is at least a little challenging. I often get really n00bish questions and that is when I start reconsidering my stance on helping people.
It depends :)
I would invest effort proportionally to the potential benefit for me (in terms of number/quality of publications) as well as the fun I would have doing it. Where the limit lies depends only on you and I can't really give you any advice on it. You should also consider other factors, such as how your own research is progressing. If you have an important deadline coming up, you shouldn't be doing something unrelated, even if it might lead to a good publication or be a lot of fun.
In any case, you should be very upfront about the level of commitment. Leaving them hanging half way is IMHO a bad idea. Not only in terms of you not being able to participate in any successes after that, but also in terms of letting someone down who asked you to help them. If you don't particularly like the project or don't want to invest any effort in it, tell them.
That said, I've been involved with a few such projects and always found it very rewarding.