Technology has always introduced changes in the way researchers administer surveys. A new technology known as virtual worlds has now emerged that promises to change data collection once again. Virtual worlds are persistent, online, computer-rendered spaces populated by hundreds, thousands, or even millions of people at a time. Previously, this population has only been surveyed in ways that required respondents to exit the virtual world before giving their answers. No survey method has existed whereby they could be surveyed while remaining present in the virtual space. Needless to say, this is less than ideal for any survey about the respondent's attitudes, perceptions, and behavior within the virtual world itself. This study introduces a method for solving this problem and a tool that allows surveys entirely within a virtual environment. The method is introduced as Virtual Assisted Self Interview (VASI), and the tool for implementing it, the Virtual Data Collection Interface (VDCI). The tool was created and deployed in the virtual world Second Life (SL), where users were asked questions about demographics and quality of life. The valid response numbers for the survey (N=2094) make it the largest in-virtualworld data collection seen so far. This paper discusses the VDCI and describes several different sampling methods, as well as results that provide unique, new insights into virtual world populations. It is found, for example, that the demographic make-up of SL is unlike that of other virtual worlds. Moreover, the SL population is unlike that of other worlds in its approach to gender-switching. The limitations and new hazards of virtual world survey research are also discussed, especially survey hacking by individuals hoping to exploit the survey for financial gain. Despite the challenges, the results generally suggest that the VDCI is a valuable new research tool for obtaining representative data on virtual world population.