The project „AI-based voice assistance for older adults with and without intellectual disabilities: ecologically valid methods applying automated speech recognition and emotion analysis (AI-Aging)” explores how older adults use an AI-based voice assistant, how the use is experienced emotionally, and which potentials and challenges this creates for the perceived quality of life and social participation. We concentrate on two user groups that previously have not found much attention in the area, although an AI-based voice assistance may be particularly useful: older adults with and without intellectual disabilities.

Our project stands out for three novel features:

• First, the use of AI-based voice assistants is investigated “ecologically valid” in the homes of older adults with and without intellectual disabilities. Despite a growing body of research focusing on the role of end-users in the context of aging and technology, there are only few empirical studies so far using this kind of approach.
• Second, the application of automated emotion analysis represents the application of cutting-edge technology, applied to our knowledge for the first time in the voice assistance context.
• Third, especially older people with intellectual disabilities have rarely been studied so far in the realm of aging, technology and digital media, although AI-based systems might come with so far untested potential.

Four-week field studies constitute the methodological key element of this project. In these field studies, older adults with and without intellectual disabilities have the opportunity to use an AI-based voice assistant in their daily life. Besides the automatically recorded data on users’ verbal and facial expressions while interacting with the system, participants are asked to document their experiences in a voice-based diary-app.

Data are subsequently analyzed using a mix of quantitative and qualitative strategies in order to generate a multi-level and multi-faceted picture of the usage of the voice assistant and its emotional-communicative implications.

In preparation for the field study, we will conduct a series of semi-structured interviews and focus group discussions. This user-centered research aims to explore in detail the needs, expectations, concerns and possible use cases of the special user groups.

We hope to contribute with our findings to identifying the incremental value of AI-based voice assistants for various groups of older adults in differentiated ways, e.g., focusing on perceived quality of life and social participation, thereby deducing recommendations for the use of AI-based technology in heterogeneous groups of older adults.

Funding

The project is funded by the Baden-Württemberg Stiftung as part of the research program "Responsible Artificial Intelligence" for a three-year period.