Non-experimental Research

Why would a researcher choose to conduct non-experimental research? What are the strengths and weaknesses associated with this type of research? Can cause and effect be established via this type of research? Why or why not?