Two experimental tasks were designed to test use of multiple-button mice. In the first, number of errors made and time to complete subtasks were measured as subjects attempted to depress one, two, or three buttons under three sets of conditions. In the second, subjects were asked to indicate true or false either by pressing one of two different buttons or by clicking a single button one or two times. People tended to be faster and more accurate using different buttons than different numbers of clicks.

The problem of finding information in on-line technical documentation is addressed by an interactive computer system called Thumb. Users access information free from the strictures of linear text, simple indices, and page numbers. Retrieval is based on a data structure roughly analogous to a detailed table of contents and a heavily cross-referenced index. The model is created and revised by an expert (such as the author) thoroughly familiar with the represented material. The expert's tasks are made easy by supportive utilities. A prototype implementation of Thumb is described and examples of dialogues with the system are given.

Current interactive programs usually provide some form of online documentation in addition to the traditional hard-copy user's manual. To save the expense of writing two documents covering the same material, it is not uncommon to find offline manuals that are available interactively as well as printed versions of material originally organized for online use. Because of the difficulties inherent in using the same material in different ways, neither approach is totally satisfactory. The THUMB system minimizes these problems by structuring offline documentation for interactive use. An expert on a particular text (e.g., its author) prepares a detailed representation of the organization of material within the document. Once this data structure (which resembles a thorough table of contents and heavily cross-referenced index) is available, users access information free from the strictures of linear text, simple indices, and page numbers. The expert's task is nontrivial, but it requires less effort than writing a new document. Creation and revision of text are made easy by supportive utilities. THUMB monitors reader's requests in order to provide experts with feedback about a document's use. Readers need not be aware of THUMB's underlying data structure or the tools available for experts.