Real-Time Assessment of Causal Attribution Shift and Stay Between Two Successive Tests of Movement Aids.

The development of welfare assistive devices for frail elderly people has attracted significant attention for its effort to improve the quality of life and reduce the burden on caregivers. However, it is challenging to conduct multiple user tests because of the significant burden on the elderly; thus, we need efficient ways to extract insight through different approaches. In this study, we aim to elucidate real-time transitions in users' emotions and achievement motivation while using such a device. We synthesize an utterance analysis method based on attribution theory, in which all users' utterances are attributed to four categories (ability, effort, task difficulty, and luck) that follow the developed coding rules. Knowing the transitions in causal attribution allows us to extract salient experiences for users, especially by extracting shifts from them and analyzing why the shift occurred and what exactly was happening before and after the shift. If only salient user experiences can be referenced from the aggregate data, useful information can be extracted in a short time to improve system characteristics and the environment. We discussed the validity and reliability of the proposed method by conducting a user test of an electric-assisted four-wheeled cycle for frail elderly people in Kakegawa city in Shizuoka, Japan. We also succeeded in marking the points that need attention, which is about 33% of the total amount of utterance data (1626 utterances), and thus confirmed the potential of the proposed method. Future research should examine how the developed methodology can help designers improve assistive device development, as well as how it can contribute to other fields such as education and social assistance.