Abstract (eng)
This master’s thesis is concerned with the phenomenon of social interaction and aims to investigate conceptualisations of social interaction within cognitive science. It furthermore investigates a specific case of human-computer interaction as a social kind of social interaction; the interaction between a human learner and a type of educational technology called ‘Intelligent Tutoring System’ (ITS). Concretely, the following three research questions (RQ) are addressed: RQ1: What conceptions of social interaction currently exist in classical/computational and situated/4E cognitive science paradigms?; RQ2: How do power and related affect influence and shape the social interaction with ITS regarding: i) social roles; ii) interaction and the (situational) structures it is embedded in; and iii) social relations? ; RQ3: Which of the conceptions from RQ1 are relevant for and can be fruitfully applied to the case of social interaction with ITS? This master’s thesis comprises three parts; one for each RQ. Part I corresponds to RQ1, Part II to RQ2, and Part III to RQ3. Part I reviews major paradigm shifts and turns within the field. Social cognition and the issues of methodological individualism, body-social problem, and human-centredness are introduced. Social interaction and related concepts across cognitive science disciples are describes, followed by a conceptual investigation of social interaction across cognitive science paradigms. Classical paradigms (computational paradigms) and situated approaches (situated/4EA cognition) are be examined, whereby the focus of Part I lies on situated cognition paradigms. Part II encompasses a comprehensive reporting of a research design proposal for addressing RQ2. The field of Intelligent Tutoring Systems (ITS) introduced. A theoretical review of social science conceptions and definition of power and affect is provided. The concrete methodological approach is reported. This includes the concrete research setting and how the ITS AutoTutor was used to create tutoring sessions to be investigated, the concrete design and content of these sessions (i.e. the curriculum and learning goals), as well as methods for data collection and analysis, which comprise a combination of Constructivist Grounded Theory and Situational Analysis. Part III addresses RQ3. This comprises a conceptual discussion of social interaction with ITS from an embodied, embedded, extended, distributed, and enactive cognition perspective. Finally, conclusions across Part I and II are drawn, including limitations and future work.