( Weber 1999): (i) How to find representative delegates of the target group(s) for participation in the design process? (ii) Does the tele-communication system optimally recognise all relevant factors that have an influence on the users' willingness and ability to use the system for tele-communication purposes? (iii) Is the system able to be easily adapted or to adapt to changes in the organisational structure or to individual characteristics during runtime? and (iv) Is the system efficient? These questions will be picked up as central themes of the next section.
Participation of members of the prospected target group is crucial for system design. At the very beginning, a case study was launched to analyse the communication behaviour of (some) BLISS users. Three user and their carers were involved. The results of interviews proved that BLISS users communicate only with a few persons. Impatience of potential communication partners or a much higher interest in BLISS (as a language) instead than in the message itself often had frustrating effects on BLISS users, so they restricted themselves to a small group of persons to communicate with. Even this small sample indicated that a much wider participation of BLISS users is necessary to recognise the variety of needs. As a conclusion, a testing environment is necessary that can be used by distributed, most often immobile participants.
Next, considerations about relevant factors that might hinder persons to use a BLISS tele-communication system are necessary. First, if the system does not sufficiently support communication or is not user-friendly (e.g., poor ergonomics), acceptance is not likely. Individual characteristics of the users, e.g., physiological, psychological or cognitive aspects as well as qualification need also intensive inspection. Furthermore, organisational characteristics of prospective contexts of use, e.g., privacy needs, required resources or physical environments, influence the design of a tele-communication system.
Adaptability due to changing needs at runtime is another important topic. E.g., users who are in the process of learning BLISS have a steadily growing vocabulary which has to be mirrored by the communication system. Therefore, individual BLISS boards are necessary that can be easily adapted and expanded. Also, if the social environment of a user changes new topics of interest may arise. On the other side, individual strategies to select symbols during sentence composition should be supported. Whereas speaking users, e.g., prefer a semantic tree structure starting
-804-
