I wouldn't belong to a society that would have me for a member... Ok, yes I would.
I have in the past been a member of the Visual Sciences Society where I presented a poster in 2006 (No relation to my previous advisor - Shapiro is a common last name, but we did put "no relation" on the poster - people found it very amusing, though they agreed it was important information). I have considered a membership in both the ACM, as well as the Human Factors and Ergonomics Society. I have also considered membership in the American Psychological Association (APA). I am currently qualified to be an Associate.
Sadly, ACM no longer appears to be offering a free alarm clock or calculator with membership, but perhaps they will again next fall. Either way, student membership is only $20, there is a local chapter at Penn State's campus, and they do have fun speakers. I went to one of their meetings last spring, and they had a speaker from the video game industry come and talk about some of the issues with AI in gaming, and careers in gaming. It was a good blend of educational content and fun graphics from game development. Also, free pizza. Really, I'm surprised more grad students don't go. Once you've graduated, it offers you a way to continue to access the ACM's library and to get in touch with other technologically savvy individuals, both in academia and in industry through the forums. This could potentially be useful.
The American Psychological Association gives accreditation to doctoral level psychology programs in the fields of school, clinical, and counseling psychology, as opposed to research only based programs. However, it is still a potentially valuable resource. Benefits include some fulltext access, and reduced rates on some journals and conferences. At $207 a year for an Associate (after the low introductory rate of $69). Student affiliate rates are only $52. But, as an affiliate, you cannot vote, and you can only vote as an Associate after 5 years of membership.
HFES: The Society furthers serious consideration of knowledge about the assignment of appropriate functions for humans and machines, whether people serve as operators, maintainers, or users in the system. And, it advocates systematic use of such knowledge to achieve compatibility in the design of interactive systems of people, machines, and environments to ensure their effectiveness, safety, and ease of performance. HFES also has a conference every year, and many IST graduate students and professors submit material.