With rapid transfer of many forms of social inquiry through structured questionnaires to the Web it is increasingly important to explore whether the Web is indeed a 'socially neutral' research tool as many believe. Because of the graphical and interactive nature of the Web and the context of global environment, social desirability effects in Web surveys may be different from with other self-administered methods, which usually reduce them. In addition, increased use of interactive services, such as multiple user domains, interactive chat rooms and interactive online games encourages widespread adoption of 'virtual personas' on the Web. It is thus important to explore how participation in such interactive services may mitigate potential benefits of the Web for social research. Our research explores whether those who are frequent participants in so-called 'alternate realities' on the Web are more likely to present their 'virtual personas' or their 'real selves' when answering questions in Web surveys. Users of interactive services are identified in a large national Web survey of Internet users in Slovenia within the project RIS (http://www.ris.org) at the Faculty of Social Sciences, University of Ljubljana. They are asked a variety of questions relating to self-image that are known to be subject to social desirability bias. At the end of the survey they are asked for their telephone number. A random sample of respondents from the population of non-users of interactive services is also selected. Both groups are then administered a telephone survey, with the key self-presentation and social desirability items replicated. We then compare the responses to the telephone survey with those provided in the Web survey. Our hypothesis is that those who are regular participants in interactive services are more likely to present themselves in a different light on the Web than on the telephone, relative to the non-user group.