Benefits
Gives Encouragement
Doing projects (research or otherwise) alone is possible; but it requires huge amount of focus and determination; especially when it comes to doing less pleasant tasks like experimentations and simulations.
From collaborations, even boring aspects of research become more pleasant. Enthusiasm for research increases. It is also more pleasurable. See benefits of pair programming in ’programming strategy’.
Toughness of problems solved
With different viewpoints, the probability of the problem being solved is higher. Also more interesting problems often discovered.
It is also a source of new problems to attack.
Papers involving collaborations tend to have much higher impact and are more cited.
Between-institution collaborations had a higher average impact.
Identifying good collaborations
If a collaboration yields much more pain than pleasure, more discouragement than encouragement, it is probably a bad collaboration. In a bad collaboration, stress is high and productivity is low.
Beginning collaborations
Present many papers. You will hear/ give ideas unrelated to what you/ others were thinking. Happened in the case of the use of affiliation networks, microscopic node analysis in solving link prediction problems; and in case of use of collaborative filtering in link prediction.
By presenting papers and generating ideas in others, reap goodwill and increase reputation.
Executing a good collaboration
Contributing
Bring in enthusiasm, ideas and work. Show ownership, take initiative.
Be prompt in accomplishing what you promised: without it, collaborators loose patience and interest, research projects then fizzle out. So, you cannot work on too many things simultaneously.
Have clear communication, supply updates upon major milestones, inform others about tough roadblocks which you could not tackle autonomously.
Feedback and setting the record
Use feedback well, but No groveling - correct unfounded statements and suppositions about your work without fail.
Getting updates
Ask if the person who is most actively involved has new thoughts/ tried new avenues about his research project.
Fix concrete date and time to discuss project, papers.