Adaptive Lernunterstützung zur interaktiven Systemnutzung für ältere Benutzer
Zusammenfassung der Projektergebnisse
In our studies, we were able to show that older people can learn very well to interact with new digital devices. Our main finding is that through appropriate user interface design and close integration of training and instruction into the interaction process, differences between young and old users can be reduced tremendously and in some cases even completely eliminated. The theoretical foundation for our experiments is the SOC-Theory, in which the acronym SOC stands for Selection, Optimization and Compensation. In an extension of this theory we shifted the interpersonal SOC-processes to extra-personal entities i.e. to computer programs. We performed seven experiments in which we varied several variables, which are assumed to support users’ learning. We reduced training time stepwise (from 90 min to less than 3 min) and changed the user interface from a traditional function-oriented/menu-based style to a task oriented/wizard-based style. The following findings can be derived from the series of seven experiments: 1) The influence of age on performance disappears when computer literacy is taken into account as a control variable. 2) In the literature, there are many recommendations for the layout of training procedures for elderly users. It is possible to develop a well-functioning and effective training program by following these recommendations. We called this "basic training". We were able to show, that elderly people can profit from basic training to a great extent. However, the already very positive results of basic training can still be improved when Guided Error Training components are included. Older users are sometimes worried about breaking a program or causing unknown system states and Guided Error Training offers a good prophylactic measure. 3) Training effects can be further improved by integrating ‘‘how’’ and ‘‘why’’ issues into the procedure. Usually, elderly people tend to learn only action sequences (step by step procedures). They reach a better understanding and interaction performance when it is explained to them how action steps are related to each other. 4) Contrary to our expectations, the introduction of role models (video-based elderly instructors) did not lead to further improvements. However, a very short instructional video (2:37 min) showing only a hand and providing some basic spoken explanations can reduce performance differences between old and young subjects. 5) With a redesign of the user interface (wizard) the differences between young and old disappeared completely; it is possible to improve the performance of older users significantly without impairing performance of young users. In all experiments, both young and old participants benefited from training and design changes, yet the gain for older users was always disproportionately greater than the gain for young users. Old and young people could perform equally well regarding effectiveness and efficiency in terms of interaction steps. Regarding time on task, older users showed significant improvements but did not reach the speed of young users.
