No Joke: An Embodied Conversational Agent Greeting Older Adults with Humour or a Smile Unrelated to Initial Acceptance

Overview

• This study examined how older adults respond to different emotional expressions in embodied conversational agents (ECAs).
• The researchers created three versions of an ECA with different facial expressions: laughing, smiling, and neutral.
• They wanted to see how these first impressions affected the older adults' acceptance and attitudes towards the ECA.

Plain English Explanation

Embodied conversational agents (ECAs) are digital assistants or companions that can communicate with people through speech, gestures, and facial expressions. As more ECAs are being developed for older adults, it's important to understand how this age group responds to them.

The researchers in this study were curious about how the initial emotional expression of an ECA might influence older adults' impressions and acceptance of it. They created three different versions of an ECA - one with a laughing expression, one with a smiling expression, and one with a neutral expression.

The researchers then had 249 older adults interact with one of the three ECA versions. They wanted to see if the positive expressions (laughing and smiling) would lead to more positive first impressions and greater acceptance compared to the neutral expression.

The results were surprising - they didn't find any significant differences in the older adults' reactions to the three ECA versions. The only factor that seemed to matter was the older adults' general attitudes towards technology and intelligent agents. This suggests that using laughter, jokes, or smiles may not be an effective way to initially engage older adults with ECAs.

Technical Explanation

The researchers used the circumplex model of affect to design three versions of an ECA with different facial expressions - laughing, smiling, and neutral. They then conducted an experiment with 249 older adult participants, randomly assigning them to interact with one of the three ECA versions.

After the interaction, the participants completed questionnaires to measure their impressions, attitudes, and acceptance of the ECA. The researchers analyzed the results to see if the facial expressions influenced these outcomes.

Surprisingly, the results showed no statistically significant differences between the three ECA versions. The only factor that appeared to matter was the participants' general attitudes towards technology and intelligent agents. This suggests that using positive emotional expressions may not be an effective way to engage older adults with ECAs initially.

Critical Analysis

The researchers acknowledge several limitations of their study. First, the interactions with the ECAs were relatively brief, so the initial impressions may not fully reflect how older adults would respond over longer-term interactions. Additionally, the study only examined static facial expressions, rather than more dynamic emotional expressions.

Another potential issue is that the study focused solely on first impressions, without investigating how the ECA expressions might impact engagement or rapport-building over multiple interactions. It's possible that positive expressions could become more effective as older adults become more familiar and comfortable with the ECA.

Overall, this study raises interesting questions about the role of emotional expression in designing ECAs for older adults. While the results challenge the assumption that positive expressions will automatically lead to better acceptance, further research is needed to fully understand how to design ECAs that effectively engage this population.

Conclusion

This study examined how older adults respond to different emotional expressions in embodied conversational agents (ECAs). Contrary to expectations, the researchers found that the ECA's facial expression (laughing, smiling, or neutral) did not significantly impact older adults' impressions or acceptance of the agent. The only factor that seemed to matter was the older adults' general attitudes towards technology and intelligent agents.

These findings suggest that using positive emotional expressions like laughter or smiles may not be an effective way to initially engage older adults with ECAs. Further research is needed to understand how to design ECAs that can effectively build rapport and foster long-term engagement with older adult users.