This paper addresses the issue of how to design online help that will really prove effective, accessible, and usable for all categories of users in the coming Information Society and, most of all, that will actually be used by novice users. The paper demonstrates the intrinsic necessity of online help and the actual failure of approaches claiming that "transparent" user interfaces eliminate the need for online support chiefly on the grounds that they encourage exploration. Empirical results in the literature or stemming from analyses of data we collected are putforward in the discussion. Based on a brief survey of the relevant literature, the major specific design issues that designers of online help systems are confronted with are presented, existing design approaches that might contribute to solving these issues are discussed, and a realistic short-term approach for improving the accessibility, effectiveness, and usability of online help, rather than the implementation of dynamic adaptation to the current user's cognitive profile or the development of contextual help systems that generate the information content of help messages dynamically according to the user's current intention and goal. We assume that it is possible, within the framework of universal design principles, to significantly enhance the effectiveness and usability of standard noncontextual help systems, mainly by making the most of the recent advances in research on multimodal interaction, especially on the integration of speech into input modalities.
展开▼
