Task orientation and user‐oriented dialog design

Work psychologists have introduced a concept, in which dialog design is a part of task design. A set of criteria of user‐oriented dialog design is presented here. These criteria are consistently integrated into a control concept. Empirical investigations on some of these criteria are then described. In an experiment to prove the criterion transparency, a desktop interface (high transparency) and a conventional menu selection interface (low transparency) were compared. The main result is the clear superiority of the user interface with direct manipulation over the conventional user interface with menu selection. Support is another of the criteria of user‐oriented dialog design that was empirically investigated. The advantages of goal‐ and task‐oriented help messages are also described here. To make the criterion flexibility and the criterion user‐definability practicable, the implementation of a dialog handler for user‐tailorable systems is introduced. The criterion participation was investigated in field studies. It will be shown that in projects with active participation the costs were exceeded to a lesser degree than in projects with passive participation and to a much lesser degree than in projects without participation.