For the past six years this research team has been testing a model for estimating the size of uncontrollable populations (populations like victims of rape, or people who are HIV-sero- positive, or homeless people). The model has shown promise but has also shown some clear shortcomings, as a result of having almost no theory regarding the rules governing social networks - that is, the rules governing who people know and what they know about one another. The investigator's program of research will a) illuminate those rules and b) improve model for estimating the size of uncountable populations. Two studies are planned. The first study addresses the question: How difficult is it to know various things about other people? Given, for example, that it is easier to know the occupation of those identified in a network than it is to know their blood type, how can this difference be measured, and what conditions of "knowing" one another does this difficulty depend on? The second study addresses the question: How many people do we know well enough to know their occupation or blood type - or a host of other kinds of information? The importance of the topic area is very great, both for fundamental research and for applied and policy purposes.
