Writing a thesis can be quite tedious and tiring – but writing a bachelor thesis within the AI-Bility project was not like that at all.
The practical foundation of my bachelor thesis was the qualitative data collected in interviews with 45 German schoolchildren. The theoretical foundation was the research article “On the Design of and Interaction with Conversational Agents: An Organizing and Assessing Review of Human-Computer Interaction Research” by Diederich et al (2022). The authors identified categories and additional constructs concerning perceptions of conversational agents and interaction outcomes. Within my bachelor thesis, I customised these concepts and constructs through complementing literature and created a frame of ten concepts and 34 constructs.
Armin, Katrin, Kilian and I transcribed 6h of interviews and then coded each transcription twice, to analyse which of the ten concepts and 34 constructs can be found in the schoolchildren’s comments.
The findings of my bachelor thesis indicated that the majority of concepts and constructs concerning children’s attitude toward and interaction with conversational agents can be found in the children’s comments.
The findings illustrated that children interact instinctively and intuitively with conversational agents. The children investigated the capabilities and functions of available conversational agents to find answers for homework or out of curiosity and favoured the opportunity to interact through natural language with conversational agents.
Critique of the children indicated contentment on an adequate level with conversational agents. Contentment attained a lower level in cases of more challenging questions for instance completing homework and contentment at a higher level in cases of uncomplicated commands and interactions
The children implied finding Cozmo and NAO more exciting and likeable compared to Alexa, Siri and Google Assistant but were even more excited about the applicability and competencies of conversational agents in film, series and social media.
The children’s opinions concerning anthropomorphism, competence, intelligence, humanness, reliability and uncanniness of the conversational agents varied highly. Though the children enjoyed interactions with friends more than interactions with available conversational agents, the children illustrated curiosity concerning conversational agents and their applicability (in days to come).
“I think one day, conversational agents will have answers to all questions and all answers will be correct!”
Quotation from one of the children