Overall... to answer these questions, one great place to start is: Couper, Mick P. (2000) "Web Surveys: A Review of Issues and Approaches" Public Opinion Quarterly. V. 64:464-494. It includes over 60 references to other papers/presentations that cover the topic. But more specifically...
1) What are the differences between paper & pencil methods versus web (including but not limited to cost, design, response, etc) I was just involved in a large scale mode experiment between web and mail. Results are to be presented at next weeks "Higher Education Center and the U.S. Department's 15th Annual National Meeting on Alcohol, Other Drug, and Violence Prevention" Conference in DC (http://www.edc.org/hec/natl/2001/). For anyone interested, I would be happy to share our slides after the presentation is given. We will also be presenting results from this same study as an invited paper at the 2002 International Conference on Improving Surveys (http://www.sfi.dk/).
2) What, if any, research exists regarding best practices? Not a lot yet... considering. There are a lot of "statements of best practices" and other non scientifically based standards. However, the arena of web survey methodology has just begun to be explored seriously. Next year at this time, I would expect to see a considerably more stable standard practice guide out there. But good quality research in methods takes time. Anyone asserting that they have "figured out" web research at this point of the game doesn't understand all of the implications.
3) Surprises, anecdotes, and special skills required to become a seasoned 'net researcher. I'd be happy to talk and share my thoughts. From my perspective... it takes a good amount of research skills... coupled with technology saavy... coupled with a general understanding of sociology, psychology, communications.
********************************** Scott Crawford Research Director - Social Science Research http://www.ms-interactive.com 734/542-7796 734/661-0323 (fax)