An empirical study of online help design: features and principles

Designers of on-line help systems have two sets of resources at their disposal: the set of features implemented in currently available systems (which are rapidly becoming a defacto standard), and a set of theoretical principles suggested by researchers in the area. There is no published evidence that either these features or principles have been empirically tested for their suitability from the users' perspective. This paper reports on an empirical study which aimed to assess the usability of a set of on-line help features and principles, in the context of users performing real application tasks. The results reveal that the more general principles associated with understandability are considered the most relevant, and that while users may complain about the design of existing on-line help features, they tend to value them more than features with which they are unfamiliar. A follow-up study showed that only minor changes need to be made to the existing defacto standard for users' concerns to be addressed, without sacrificing the advantages of familiarity. The study addresses questions of context sensitivity, obtrusiveness and the importance of definitions, and highlights the usefulness of questioning emerging defacto standards that have not been based on empirical studies