Long ago, I wrote a paper called How to Succeed in Graduate School: A Guide for Students and Advisors, that has been widely distributed on the Internet. (The original version of the paper was called How to Be a Good Graduate Student / Advisor.) It contains a lot of suggestions on finding an advisor and a thesis topic, doing research, writing the thesis, and establishing a research network, among other things.
The latest version of the paper is available in postscript and PDF forms. There are some slides that summarize some of the key points in the paper, in postscript form and in the original latex version.
The paper has been published in two parts in issues 1.2 and 1.3 of Crossroads, the online ACM student magazine.
The paper has also been translated into Chinese.
An HTML version of the original paper can be found at Indiana University's What Every New Grad Student Should Know. This page also points to Phil Agre's networking paper and other useful resources. Another HTML version is at UMBC's Information for Graduate Students, another useful page.
A useful collection of pointers to online resources and advice for graduate students is the OpenDirectory Project's "How to Study" page on PostGrad Research. You may also be interested in another website, Getting In: An Applicant's Guide to Graduate School Admissions, by Dave Burrell. A mailing list that may be of interest to female graduate students is the systers-students mailing list.
The paper was also reprinted in the IAPPP (International Amateur-Professional Photoelectric Photometry) Communications, and in a shorter form in Vivek, an India-based quarterly in AI.
Anyone who would like to distribute or excerpt the paper is welcome to do so, as long as my name remains attached, and proper attribution is given to me.
Some of the references in the paper are incomplete (or possibly incorrect). If anyone has more complete bibligraphic information for any of the references, I'd appreciate it if you would send it to me. Comments and feedback on the paper are also very welcome.My user ID is mariedj, and the machine name of my e-mail address is cs.umbc.edu. Human beings should be able to compose an e-mail address from this information, but hopefully, web crawlers won't!