Resumen
Thousands of social gatherings - online communities ? occur via ICTs across the Internet. They use listservers, bulletin boards, instant messaging, blogs, social network and meeting facilitation systems and purpose-build platforms. Each community is the product of a complex set of interacting variables. Some of these variables come from usability of the software but social factors have a strong influence. Direct cause and effect relationships are often difficult to tease out and identifying criteria for successful online communities can be difficult. Furthermore several disciplines own this topic, each with its own theories, research methods, community of researchers and publication outlets. This is good for multidisciplinary research but it can be challenging. In this paper I discuss two research projects. Each focuses on community populations that are hard to reach. The first analyses why people do not actively participate in online discussions, i.e., lurk. The second is a new project to develop cross-cultural online book communities for children across the world who do not speak each other?s languages. I present the key findings from each project and suggest ways of working with these hard to reach populations.