Methods for defining user groups and user-adjusted information structures

A common problem in the design of information systems is how to structure t he information in a way that is most useful to different groups of users. T his paper describes some statistical methods for revealing the structure in herent in empirical data elicited from users. It is illustrated by the appl ication of these methods to the design of some web pages giving information about the Universitat de Valencia. Three potential user groups were identi fied, administrative staff, teaching staff and students. The first analysis demonstrated that users within these three groups assign relatively homoge neous structures, but that the structures assigned by the three groups are not the same, and also, teaching and administrative staff were shown to be relatively similar and different from students. Second, the ideal informati on structures for each group were identified and validated against the orig inal data. The methods described can be applied to any design situation whe re there is an existing user population that can be called on to provide da ta.