Thursday 8th of May 11.15 – 12.45
Aula Volta
Chair of the session – Paola Previtali
| 11.15 – 11.35 | Benjamin Stodt | Detection of changes in sound locations in real and virtual environments: Does age affect perception an neural processing? |
| 11.35-11.55 | Burcu Demiray | Older adults’ learning interests, needs and challenges with AI tools. |
| 11.55 – 12.15 | Sonja Pedell | Stimulating and fun use of generative AI tools by older adults in a retirement village for producing digital stories. |
| 12.15-12.35 | Mario Bonato | Auto-global examination of Mental State (Auto-Gems): a web-based cognitive screening. |
Abstracts
Detection of changes in sound locations in real and virtual environments: Does age affect perception an neural processing?
by Benjamin Stodt1 | Daniel Neudek2 | Rainer Martin2 | Stephan Getzmann1
1 Leibniz Research Centre for Working Environment and Human Factor (IfADo), Dortmund, Germany
2 Institute of Communication Acoustics, Ruhr- Universität Bochum, Bochum, Germany
With an aging workforce, understanding how aging influences routine sensory-cognitive processes, like audio-visual perception in everyday situations, is becoming increasingly important. Virtual reality (VR) facilitates research in ecologically valid contexts. However, it is essential to ensure that findings obtained in VR align with those observed in real-world environments. 22 younger (M=24.00 years) and 22 older participants (M=64.77 years) performed an auditory change detection task in two environments: a real laboratory room with physical loudspeakers and a virtual twin of this room (presented via head-mounted display and headphones). Participants had to respond whenever the sound position shifted to a target loudspeaker positioned in front, behind, to the left, or to the right relative to a central (standard) loudspeaker. During this, behavioral and neurophysiological data (EEG) were collected.
In general, comparable results were observed in both environments, validating VR as a suitable tool for auditory tasks in more life-like scenarios. In both settings, expected event related potentials representing neurocognitive correlates of deviance detection (mismatch negativity) and attentional orientation (P3b) were identified. However, the characteristics of these potentials partially varied depending on the environment, target position, and age group. In a follow-up study, we further explored the influence of additional visual target cues among older participants.
Older adults’ learning interests, needs and challenges with AI tools.
by Lea Riniker1 | Kathrin Inerle1, 2 | Burcu Demiray1, 2
1 Department of Psychology, University of Zurich
2 Healthy Longevity Center, University of Zurich
In two studies, we examined older adults’ learning interests, needs and challenges with AI tools. In the first online survey, we asked about experiences with ChatGPT. Our preliminary results (sample mean age 71.2; data collection ongoing) show that 58% have used ChatGPT and 81% would like to learn more about it. Their challenges included data security concerns, lack of efficient prompts and uncertainty about benefits. In the second online survey, we examined how avatar instructors in e-learning videos are perceived. A total of 146 adults (mean age 69.6) were randomly assigned to an experimental or a control condition. Control group watched an e-learning video with a human instructor, whereas the experimental group viewed the same content taught by an avatar. All participants took a quiz and evaluated the instructors. 54% of the avatar group reported noticing that the instructor was an avatar. The use of an avatar did not affect performance, as the two groups achieved the same quiz score. The human instructor was rated only slightly higher than the avatar in terms of gestures, sympathy, authenticity, and trustworthiness. Both instructors were perceived as similarly competent. Results are discussed in line with the importance and digital transformation of lifelong learning.
Stimulating and fun use of generative AI tools by older adults in a retirement village for producing digital stories.
by Sonja Pedell | Diego Muñoz | Lan An
Swinburne Living Lab, Swinburne University of Technology, Melbourne Australia
Generative AI (GenAI) has much potential, but many older adults are unsure how to gain value from it. Our research investigated the use of GenAI tools by older adults living in a retirement village for creating digital stories. The participants used GenAI for different purposes: generating images (Craiyon and DeepAI Image Generator) to depict dear memories in an exact manner, creating narratives (speech-to-text and text-to-speech generating different accents such as Scottish voices), and creating fun avatars with Bitmoji. We co-explored using these tools, with residents then trying them by themselves. Our ethnographic data from the one-year weekly engagement with five residents showed that using these GenAI tools for generating story content happened in creative and playful ways enabling them to decide if, how, and why to use the GenAI tools. Exploring GenAI applications, residents broadened the use of their own devices by learning about practical uses of AI, catering for personalised needs and develop new skills. Based on the data collected during the production process, residents talked competently and confidently about GenAI, justifying clear choices about when and how to use it to create an authentic story to convey to their audience. The findings suggest that teaching novel AI tools to older adults can be barrier-free, while being stimulating in offering a wide variety of image styles and audio choices. Moreover, group members were sharing unexpected results which often caused laughter and lively discussions. Older adults understood the technology and used the GenAI tools comfortably being able to explain the use to a broader audience upon completion of their stories.
Auto-global examination of Mental State (Auto-Gems): a web-based cognitive screening.
by Mario Bonato1 | Veronica Pucci1 | Giulio Contemori1 | Maria Silvia Saccani1, 2 | Giorgio Arcara2 | Sara Mondini1, 2
1 University of Padova
2 IRCCS San Camillo, Lido Venice
Recent methodological developments have contributed to a significant advance in computerised neuropsychological instruments and procedures. Here we present Auto- GEMS, a newly developed, web-based, screening test allowing to quickly estimate an individual’s cognitive state also considering their cognitive reserve. Auto-GEMS can be administered from remote and it measures cognitive functioning on eleven items similarly to the in-person paper-and-pencil version (GEMS) and to the remote (phone or video call) version (Tele-GEMS) of the same screening. We collected normative data on a sample of 1308 Italian-speaking participants (age range 18–93) to verify its psychometric properties and computed regression models on demographic variables to establish clinical cut-offs. The psychometric properties of Auto-GEMS have shown good internal consistency, test-retest reliability and convergent validity. This short and user-friendly tool has a number of potential applications. For instance, it can be useful in clinical practice to monitor the cognitive profile of patients or vulnerable individuals, or even administered in a face-to-face, standard clinical setting. It can also be used in research studies to screen participants. The testing materials and the collected data are freely available in a digital archive along with a web App to visualise the test outcome with reference to its normative data.