The rather old concept of human "exchange theory" which roxanne wrote up in "the network nation" was very good at expressing participation in discussion threads where when one wrote something the fact that they got replies encouraged them to contribute more comments. Exchange theory would also state that the reward of gaining status as an expert for the contributions made to dealing with problems would also be a coin of exchange. In networks of collaborative groups it is clear that a contribution will get payback based upon the number of other collaborators. We also found when in those days we could measure the time a person spent in a single conference or a message list that they would tend to sign on often enough to get like seven +- two new items. The more active the discussion the more frequent they would sign on. when there was much less they would wait to sign on for longer periods and this lead to the observation of critical mass needed to keep a discussion going. Also at some point information overload sets in when they cannot sign in often enough to keep up with the flow. This was in the paper that roxanne and i did some years back for cacm (i think 1981) on "information overload." In delphi one of the most important motivations to get professionals to participate is to make sure they know they are going to be collaborating with a peer group where they may get more feedback of useful information than they will have to put in. Sometimes we let participants know who is in the group even if it is anonymous with respect to entries. That is I feel a good motivation for professional message lists and when one starts getting too many requests for information verses contributions of information or too many announcements of paper requests i tend to lesson my activity in the message list. i did a study a year back on information seeking behavior of professionals in emergency management and as a result, i published recently through springer a paper on a recommender system design for communities of practice based upon some of the above areas including Delphi design. i can send it to anyone interested. The original study for NLM (on my website) was not allowed to incorporate any solutions to the information seeking behavior problems of emergency management professionals. But if you read the 60 page raw appendix of their actual text answers to our open ended questions the solutions become clear and in the paper i later published. The original inquiry was a one round delphi on information seeking behavior and problems using a snowball sampling and open ended text questions in a word file we sent people. By the way people who make comments but never get replies tend to stop making comments. Distinguished Professor Emeritus Information Systems, NJIT homepage: http://is.njit.edu/turoff